RU2405716C2 - System and process for automatic control of bearing wing aerodynamic surfaces flight - Google Patents

Abstract

FIELD: transport.

SUBSTANCE: invention relates to aircraft engineering. Proposed system 1 comprises appliances 3, 5 for onboard and ground detection, appliance 7 to process first 3a and second 5a pieces of data and convert said data in mechanical drive to make aerodynamic surface 2 of bearing wheel move along the trajectory, and system to transmit first pieces of said data into data processing and control appliance 7. Step-by-step control over bearing wheel aerodynamic flight is carried out with the help of system 1.

EFFECT: possibility to control wing aerodynamic surfaces flight in rotary system.

36 cl, 8 dwg

Description

The present invention relates to a system and a process for automatically controlling the flight of aerodynamic surfaces of a power wing, in particular for optimizing the generation of electricity by flying aerodynamic surfaces of a power wing connected to a carousel-type system.

As you know, there is an extensive literature and a significant number of technical solutions related to automatic control in flight of autonomous aircraft (UAV). As you know, the possibility that a person controls the flight of the aerodynamic surface of a wing, such as, for example, a kite, mainly arises from the assessment by seeing the position and orientation of the aerodynamic surface of the wing in space, which provides a set of perception data that provide the possibility of correction maneuver of traction cables. Automation of the maneuver of the aerodynamic surfaces of the wing inevitably takes place through the exact reproduction of this human sensitivity.

However, reference materials and literature related to this prior art do not show solutions or studies that are aimed at automatically controlling the flight of aerodynamic surfaces of a power wing, in particular, implemented as a “power kite”. In fact, it is assumed that the problems of this proper management are multiple and complex, so that they require the most appropriate application of the most advanced management methodologies and algorithms. The flight of the aerodynamic surface of the power wing and its modeling, in fact, is directed to the use of non-linear systems with many variables, while the specifications of the control functions should be analyzed using the appropriate reliability requirements for parametric changes and dynamics, which cannot be modeled with sufficient accuracy. Depending on these characteristics, the control system can also provide the exact control definition functionality designed on a virtual prototype using experimental measurements of a real system during its implementation. Problems formulated to control real systems by approximating mathematical models of the system used to design a control system have always been taken into account by researchers in the field of technology based on the main works of Nyquist and Bode. Nevertheless, only from the 70s-80s did the corresponding development of the results appear, allowing one to systematically and quantitatively address the effect of the uncertainty of the models used for the analysis and synthesis of control systems, which led to the extensive development of the field of robust control. Since these methodologies can be used to solve a significant part of real problems, it is necessary that such characteristics be obtained through appropriate identification methods that operate with measurements performed in a real system that must be controlled, called robust identification in the reference literature, oriented on identification or identification management by belonging to many. These aspects were mainly considered in the following works:

Horowitz, "Synthesis of Feedback Control Systems", Academic Press, 1963;

Menga G., Milanese M., Negro A., "Min-max quadratic cost control of systems described by approximate models", IEEE Trans. Aut. Contr, 1976;

J.C. Doyle, "Guaranteed margins for LQG regulators", IEEE Trans. Aut. Contr, 1978;

V.L. Kharitonov, "Asymptotic stability of an equilibrium position of a family of systems of linear differential equations", Differential Equations, 1979;

G. Zames, "Feedback and optimal sensitivity", IEEE Trans. Aut. Contr, 1981-1982;

H. Kimura, "Robust stabilizability for a class of transfer functions", IEEE Trans. Aut. Contr, 1984;

J.C. Doyle, K. Glover, P.P. Khargonekar, B.A. Francis, "State space solution to standard H-2 and H-inf control problems", IEEE Trans. Aut. Contr, 1989;

S.P. Bhattacharyya, H. Chapellat, L.H. Keel, "Robust Control: The Parametric Approach", Prentice Hall, 1995;

K. Zhou, J. C. Doyle, K. Glover, "Robust and Optimal Control", Prentice Hall, 1996;

M. Milanese, R. Tempo, A. Vicino (Eds), "Robustness in Identification and Control", Plenum, London, 1989;

IEEE Trans, on Aut. Contr., "Special Issue on System Identification for Robust Control Design", 1992;

A.B. Kurzhanski, V.M. Veliov (Eds), "Modeling Techniques for Uncertain Systems", Birkhauser, 1994;

B. Ninness and G. C. Goodwin, "Estimation of model quality", Automatica, 1995;

M. Milanese, J. Norton, H. Piet-Lahanier, E. Walter (Eds), "Bounding Approaches to System Identification", Plenum Press, 1996;

J. R. Partington, "Interpolation, Identification, and Sampling", Clarendon Press, 1997;

H. Kimura, M. Milanese (Org.), Invited Session "Model Set Theory in Identification and Control", 38th IEEE CDC, Phoenix, 1999;

J. Chen, G. Gu, "Control-oriented system identification: an H-infinity approach", John Wiley, 2000;

Int. J. of Robust and Nonlinear Control, Special Issue on "Robust control from data", M. Milanese, M. Taragna Eds., 2004.

In addition to the above articles and books, contributions at the international level to the development of innovative techniques and algorithms for robust identification and management topics are documented in additional international literature; in particular, methods for identifying approximable models of complex linear and nonlinear systems are considered in:

M. Milanese, G. Belforte: "Estimation theory and uncertainty intervals evaluation in presence of unknown but bounded errors: linear families of model and estimators", IEEE Transactions on Automatic Control, vol. 27, n. April 2, 1982;

M. Milanese, R. Tempo: "Optimal Algorithms Theory for robust estimation and prediction", IEEE Trans. AC, August 1985;

B.Z. Kacewicz, M. Milanese, A. Vicino: "Conditionally optimal algorithms and estimation of reduced order models" Invited paper 2nd Int. Symposium on Optimal Algorithms, New York, 1987. Also Journal of Complexity ol. 4, pp. 73-85, 1988;

M. Milanese, A. Vicino, "Optimal estimation theory for dynamic systems with set membership uncertainty: an overview", Automatica, vol. 27, 997-1009, 1991;

L. Giarre, B. Z. Kacewicz, M. Milanese, "Model quality evaluation in set membership identification", Automatica, vol. 33, no. 6, pp. 1133-1139, 1997;

M. Milanese, M. Taragna, "Optimality, approximation, and complexity in Set Membership H-inf identification", IEEE Transactions on Automatic Control, vol. AC-47 (10), pp. 1682-1690,2002;

M. Milanese, C. Novara, "Set Membership Identification of Nonlinear Systems", Automatica, Vol. 40/6, pp. 957-975, 2004;

K. Hsu, M. Claassen, C. Novara, P. Khargonekar, M. Milanese, K. Poolla, "Non-Parametric Identification of Static Nonlinearities in a General Interconnected System", International Federation Automatic Control World Conference, Prague, 2005.

Robust control starting with experimental data is reviewed in:

M. Milanese, G. Fiorio, S. Malan, "Robust performances control design for a high accuracy calibration device", Automatica, Special Issue on Robust Control, vol. 29, pp. 147-156, 1993;

S. Malan, M. Milanese, D. Regruto and M. Taragna, "Robust control from data via uncertainty model sets identification", International Journal of Robust and Nonlinear Control, Special Issue on "Robust control from data", 2004.

Robust control in cases where there are saturations with twisting prevention techniques considered in:

M. Canale, M. Milanese, "Robust design of predictive controllers in presence of unmodeled dynamics", European Journal of Control, vol. 9, no. 5, 2003;

M. Canale, M. Milanese, Z. Ahmad, E. Matta, "An Improved Semi-Active Suspension Control Strategy Using Predictive Techniques", Proc. IEEE International Conference on Information & Communication Technologies, Damasco, 2004; and their applications to different application sectors are dealt with by:

M. Milanese, C. Novara, P. Gabrielli, L. Tenneriello, "Experimental Modeling of vertical dynamics of vehicles with controlled suspensions", SAE World Congress, Detroit, Michigan, 2004;

M. Milanese, C. Novara, "Set Membership Prediction of River Flow", Systems and Control Letters, Vol. 53/1, pp. 31-39, 2004;

A. Chiesa, "Tecniche di controllo Fault Tolerant per velivoli senza pilota (UAV)" Graduating Paper, responsible M. Milanese, Politecnico di Torino, 2004;

M. Milanese, C. Novara, L. Pivano, "Structured SM identification of vehicles vertical dynamics", Mathematical and Computer Modeling of Dynamical Systems (Special Issue), 2005.

However, from what was stated above, unknown systems and / or processes for automatically controlling the flight of aerodynamic surfaces of the power wing, which work predictably, namely, depending on the observation and prediction of future flight conditions of the aerodynamic surfaces of the wing themselves, and which allow taking into account critical situations and forecasting errors.

Patent application (Italy) number TO2003A000945 and Patent application (Europe) EP 04028646.0, which claims the priority of the applicant, disclose a system for converting the kinetic energy of wind flows into electrical energy through predictive and adaptive flight control of the aerodynamic surfaces of the power wing connected to the carousel system type. In such a system, it should be desirable to use a system and process for automatically controlling the flight of the used aerodynamic surfaces of the power wing, which provide the ability to set in motion real-time aerodynamic surfaces of the wing according to the modes described in the aforementioned applications. However, in the prior art systems are not known that provide the ability to control the flight of the aerodynamic surfaces of the wing, effectively used in a system of "carousel" type.

Therefore, the aim of the present invention is to solve the problems of the prior art by providing a system and a process for automatically controlling the flight of aerodynamic surfaces of a power wing in a predictable mode according to a “preferred control strategy” based on observing and predicting future flight conditions of the aerodynamic surfaces of the wing, taking into account critical situations and errors due to forecasting, allowing to avoid local maxima, fluctuations and abilnostey management.

Another objective of the present invention is to provide a system and a process for automatically controlling the flight of aerodynamic surfaces of a power wing used in the carousel type system described in Patent Application (Italy) No. TO2003A000945 and Patent Application (Europe) No. EP 04028646.0.

The above and other objectives and advantages of the present invention, obvious from the following description, are achieved using the automatic flight control system of the aerodynamic surfaces of the power wing according to claim 1.

Moreover, the above and other objectives and advantages of the present invention are achieved using the automatic flight control process of the aerodynamic surfaces of the power wing according to claim 27.

Preferred embodiments and non-trivial variants of the present invention are the subject of the dependent claims.

The present invention should be better described by means of some of its preferred embodiments provided by way of non-limiting example, with reference to the accompanying drawings, in which:

figure 1 is a block diagram illustrating the main components of the automatic flight control system of the aerodynamic surfaces of the power wing according to the present invention;

figa is a diagram showing a reference system associated with the aerodynamic surface of the power wing, limited to the known system of "roundabout" type;

fig.2b is a graph showing a vector decomposition of a vector showing the acceleration of gravity;

figure 3 is a diagram showing the aeronautical region of the aerodynamic surface of the power wing relative to the direction and orientation of the wind flow;

figa is a diagram showing a three-dimensional flight target of the aerodynamic surface of the power wing in accordance with the process according to the present invention;

fig.4b shows a top view of the circuit of figa;

figure 5 shows the diagram of figure 4, equipped with certain process parameters according to the present invention; and

Fig.6 shows a top view of a known system of "roundabout" type at some stages of flight of the aerodynamic surface of the wing, driven by the system and / or process according to the present invention.

In the following description, the system and process according to the present invention is described as being preferably applied, simply by way of example, to the automatic flight control of aerodynamic surfaces of a power wing used according to what is described in Patent Application (Italy) No. TO2003A000945 and in Patent Application (Europe) No. EP 04028646.0. Therefore, due to obvious brevity issues, for a detailed description of the components that are mentioned above and are common with the well-known "carousel" type system, reference should be made to the above applications. However, it is completely obvious that the system and process according to the present invention can also be used for other applications other than the "carousel-type" system, by making modifications that are in any case within the understanding of those skilled in the art.

In this context, it is assumed that the flight of at least one aerodynamic surface of the power wing is controlled by a drive unit equipped with winchs alternately driven by a motor, to which the aerodynamic surface is connected via two corresponding drive cables, as described in TO2003A000945 and EP 04028646.0.

With reference to figa, it can be noted that the automatic flight control system 1 of at least one aerodynamic surface 2 of the power wing according to the present invention contains:

- the first means of detection 3 on board the aerodynamic surface 2 of the power wing, configured to detect the first pieces of information 3a related to at least one position and one orientation in the space of the aerodynamic surface itself and the three-dimensional accelerations to which it is subjected;

- second ground detecting means 5, configured to detect second pieces of information 5a related to at least one tension value of the drive cables of the wing aerodynamic surface 2 and to the position of the counterweight of the drive unit 9;

- means 7 for processing and controlling the first 3a and second 5a pieces of information, made with the possibility of converting the contents of these pieces of information into a mechanical drive, operating with the winches of the drive unit 9, to drive the aerodynamic surface 2 of the wing along the flight path, which maximizes the effect of "lift" generated on the aerodynamic surface 2 of the wing by means of the wind flow into which it is immersed, and maximizes the magnitude of the kinetic energy subtracted from the wind flow; in fact, it can be demonstrated that if the aerodynamic surface of the power wing is free from scanning the wind front in the wind flow in the “lift” mode, not only the thrust that it can apply to the drive cables (and as a result, possibly to the levers of the “carousel” system) far exceeds the thrust applied by keeping the aerodynamic surface of the wing stationary at the point of maximum wind resistance due to the use of the drag effect, but also the area in which the aerodynamic surfaces of the wing and have a braking effect on the rotation of a possible "carousel" system is completely eliminated; in particular, the processing and control means 7 comprises a geometric motor 7a adapted to process these first pieces of information 3a in order to return information 7c about the position, acceleration and orientation of the aerodynamic surface 2 of the wing to the numerical control unit 7b, essentially of a traditional type, made with the ability to operate 9a with the winches of the drive unit 9 to control the traction force on the drive cables; and

- a system for transmitting first portions of information 3a to the processing and control means 7 and, in particular, to the geometric motor 7a.

Moreover, the system 1 according to the present invention may comprise an unstable dispersion drive realized according to the robust control theory.

To better understand the modes in which the first pieces of information 3a are directly collected by the first means of detection 3, what other pieces of information can be indirectly obtained from the first pieces of information 3a, and then what types of the first means of detection 3 can be used in the system 1 according to the present invention, it should be useful to briefly analyze the geometric portion of information that characterizes the position of the aerodynamic surface 2 of the wing in space. Therefore, with reference to FIG. 2a, it can be noted that each wing aerodynamic surface 2 that extends from under the lever 20a of the carousel-type system 20 according to the applications TO2003A000945 and EP 04028646.0, describes by means of two drive cables 21 pressing it to the ground, and the imaginary line L that connects its ends, the triangle OAB, lying on a plane in space. Aerodynamics research introduces the concepts of roll, pitch, and course deviation in general position in the space of an aircraft. In the case of the aerodynamic surface 2 of the wing, there is a restriction, represented by the exhaust nozzle of the drive cables 21, from the lever 20a, which forces one to abandon the classical terminology (deviation from the course, roll, pitch). Next, consider the ideal base coordinate system XYZ _ref common with the lever 20a of the system 20, so that the acceleration of gravity has a component along only one Z axis. Let us further consider an imaginary line that connects the end of the lever 20a to the midpoint of the wing aerodynamic surface 2. This segment, described above, describes the angle α with the horizontal plane XY _ref , and the angle γ, which can be found starting from the axis X _ref , by projecting the segment onto the horizontal plane XY _ref . The angles α and γ determine the position of the aerodynamic surface 2 of the wing in space. However, the geometric portion of information is complete only when the concept of orientation is also introduced. To define it, let's look at the above triangle again. The triangle OAB lies on a plane whose position relative to the ground coordinate system varies with time depending on the flight of the aerodynamic surface 2 of the wing. The term orientation of the aerodynamic surface 2 of the wing, therefore, defines the angle β described by the plane on which the triangle and the horizontal plane Xy _ref lie, not counting the angles α and γ. Although in order to logically determine the position and orientation of the aerodynamic surface 2 of the wing in space, an artificial ground-based optical vision system can be adapted, which can constantly deteriorate due to the possible passage of clouds or the absence of a transparent atmosphere, and these problems are felt even more if system 1 is used in conjunction with a carousel-type system 20, characterized by high flight altitudes of the wing aerodynamic surface 2.

Therefore, it is preferable in order to logically determine the position and orientation of the aerodynamic surface 2 of the wing in space and the accelerations to which it is subjected, the first detection means 3 comprises MEMS-type three-dimensional accelerometers in combination with at least one electronic compass. The latter can be implemented using a flux-gate magnetometer or another type, which allows to provide a reliable solution, much more accurate than could be obtained by artificial viewing even in conditions of perfect visibility. In particular, the aerodynamic surface 2 of the wing is equipped with at least one magnetometer and at least two three-dimensional accelerometers located at the ends of the aerodynamic surface of the wing, preferably near the place where the drive cables are connected to the walls of the aerodynamic surface of the wing. Accelerometers on board the aerodynamic surface of the wing, therefore, solve the functions:

- providing a system 1 with the ability to recognize the position and orientation of the aerodynamic surface of the wing in space;

- providing multidimensional and instantaneous acceleration data useful for the correct feedback of the process control loop according to the present invention described below, implemented by system 1;

- Correction of a possible false perception of the gravity vector caused by strong aerodynamic accelerations.

Magnetometers provide only one of the pieces of information, which is not necessarily located near the accelerometer system, namely the rotation of the aerodynamic surface 2 of the wing around the axis of the center of gravity.

Accelerometers of the type used in system 1 are sensitive to a wide range of accelerations, which range from static acceleration, such as acceleration of gravity, to phenomena with frequency characteristics of several kHz.

The three-dimensional accelerometer, obviously, determines its own Cartesian coordinate system XYZ _A , as shown in fig.2b. By displaying the moment at which this coordinate system coincides with the ideal Cartesian coordinate system XYZ _ref , the sensitivity to gravity (static) acceleration allows you to de facto distinguish the change in acceleration due only to the inclination of the accelerometer (which implies a change in direction Z _A relative to Z _ref ) from the change due to the actual displacement of the origin of the coordinate system XYZ _A , defined as aerodynamic acceleration, which represents the flight of the aerodynamic surface of the wing.

In general, a three-dimensional accelerometer will have a random position in space. The vector g , which describes the acceleration of gravity with a constant modulus, direction and sign, therefore, can be decomposed into its three components along the versors parallel to the three axes X _A , Y _A , Z _A. Obviously, the position of the gravity vector g in the coordinate system XYZ _A can also be expressed in spherical coordinates depending on the angles φ and θ and the module g (9.8 m / s ² ) by the following change of coordinates:

X _gA = g * cos φ * sin θ

Y _gA = g * sin φ * cos θ

Z _gA = g * cos θ

from which the following is obtained:

φ = atan ² (X _gA , Y _gA )

θ = asin (X _gA )

where atan ² is the arc tangent function (x) with an uncertainty resolution of (+/-) π / 2.

Each accelerometer is subject to two components of acceleration. The acceleration of gravity described above is vectorially summed with aerodynamic acceleration due to the actual movement of the aerodynamic surface of the wing relative to a fixed coordinate system XYZ _ref . The first detection means 3, therefore, can be implemented with its own intelligence on board the aerodynamic surface of the wing, the corresponding algorithms, the purpose of which is to distinguish gravity from aerodynamic acceleration, transmitting, on the one hand, spherical coordinates that indicate decomposition of the gravity vector relative to the coordinate system of the accelerometer (and, therefore, the inclination of the accelerometer relative to a fixed coordinate system ), And on the other hand, the evaluation of the aerodynamic acceleration in real time. The measurement of such acceleration allows, firstly, to implement real-time control techniques that are preferred for quick control of the flight of the aerodynamic surfaces of the wing, as will be shown later in the description of the control process according to the present invention. This measurement additionally provides the ability to instantly adjust the required angles φ and θ to assess the orientation of the accelerometer, while the integration of acceleration data provides the opportunity to additionally assess the flight path of the aerodynamic wing surface, de facto completing all the information associated with this knowledge.

The need to provide the wing aerodynamic surface with at least two accelerometers stems from the fact that it is necessary to distinguish these movements of the wing aerodynamic surface, which can be considered as rotations around one of its ends. In this case, only one accelerometer, collected, for example, in the center of the aerodynamic surface of the wing, must perceive the tangential velocity, which can be approximated using υ _t = ω • r, where ω is the rotation speed of the corresponding circular motion, while r is the distance of the accelerometer from the center of rotation (in this case, half the width of the aerodynamic surface of the wing). Such a velocity does not correctly describe the movement of the aerodynamic surface of the wing, the free end of which describes a circular path with a double tangential velocity υ _t = ω • r and centripetal acceleration equal to a _c = υ _t ² / r. Two accelerometers located on the aerodynamic surface of the wing, therefore, implement an inertial platform with six axes with high performance and high cutoff frequency, which allows you to recognize movements on six axes and describe the position and orientation of the aerodynamic surface of the wing.

Rotations around an axis that do not imply changes in the direction of the axis that is parallel to Z _ref , however, can only be perceived by this configuration of accelerometers in interaction. Therefore, it is necessary to provide the aerodynamic surface of the wing with at least one magnetometer to compensate for drift due to double integration. Two accelerometers together with the exit point of the cables 21 from the lever 20a describe a fully known triangle, the length of each side of which is known. The only unknown data is the angle included between the lever and the pair of cables 21 (consider the bisector of the acute angle included between the two cables 21) projected onto the horizontal plane XY _ref . This angle can be more conveniently estimated from the ground by measuring it directly at the point where the cables exit the lever.

As mentioned, the second detection means 5 is configured to detect second pieces of information 5a related at least to the magnitude of the tension of the drive cables of the wing aerodynamic surface 2 and the position actually or imitated by the winches of the drive unit 9 of the counterweight; in particular, the counterweight function potentially or electrically absorbs and stores excess energy, which can be generated due to too strong wind, and returns it at the stages in which the aerodynamic surface of the wing is in a state of speed loss relative to the wind. The second detection means 5 may therefore comprise “strain gauges” for measuring the deflection of the drive cables and encoders on the winches of the drive unit; these encoders, possibly associated with a reciprocating motor, can also be used to detect the length of the drive cable unwound from the winches, and then the distance from the aerodynamic surface of the wing to the drive unit and the differential length between the two drive cables of the same aerodynamic surface of the wing. Moreover, the second detection means 5 may also comprise proximity sensors configured to detect the angle between the drive cables at the outlet of the nozzle of the lever of the carousel system.

The second detection means 5 may also include an optical or microwave ground-based artificial vision system for the position of the aerodynamic surface of the wing. Terrestrial optical viewing, in comparison with microwave viewing, has an enduring disadvantage of depending on the passage of clouds that hide the aerodynamic surface of the wing from view. An effective artificial vision system, however, provides an important contribution to safety, providing the necessary information in order to avoid collisions with helicopters and small aircraft in general.

From the first 3a and second 5a pieces of information, respectively measured by means of the first 3 and second 5 means of detection, the position of the aerodynamic surface of the wing in space can in any case be obtained by at least three different methods:

a) processing data from accelerometers and magnetometers through a geometric motor; in particular, the length of the position vector of the aerodynamic surface of the wing can be obtained through double integration of the acceleration signal;

b) a combination of the data that can be obtained from the winch encoders when measuring the angles between the cables and the lever that can be obtained at the end of the lever; it should be noted that from the end of the lever only angles can be estimated, which, in general, are formed by the pair of drive cables using the lever itself;

c) the use of an artificial vision system: in this case, however, a delay due to the collection and compilation of images should be added to the delay in extracting information.

Similarly, the orientation of the aerodynamic surface of the wing in space can be obtained by processing the data of the accelerometer and by means of an artificial vision system.

The acceleration, instead, must necessarily be obtained on board the aerodynamic surface of the wing due to the fact that the delay introduced by calculating the second derivative of the position is incompatible with real-time control techniques that are preferred for controlling the flight of the aerodynamic surfaces of the wing. This implies that intelligence on board the aerodynamic surface of the wing becomes an integral part of the control system 1.

In an alternative embodiment, it is also possible to provide for the use of rheological polymers in the structure of the aerodynamic surface of the wing in order to implement moving systems directly on board; in this case, it can be ensured that the first detection means comprises other sensors that can provide signals extracted by feedback from the composite materials to facilitate or detect the position of the aerodynamic surface of the wing in space.

Based on what is described above, measuring the position and orientation of the aerodynamic surface of the wing, as a result, is likely to be unnecessary; in particular, the assessment of acceleration based on information of the immediate position and orientation, although it is not effective in relation to redundancy in real time, can contribute to the formation of diagnostic redundancy of the system to assess the flight characteristics of the aerodynamic surface of the wing.

In this context, particular relevance may be given to invented methodologies for designing virtual sensors shared with the first and / or second detection means of system 1. In fact, many of the quantitative quantities measured by the sensors of the first and second detection means may, in the event of a failure of a particular sensor can also be estimated from measurements of other sensors by means of experts / Kalman filters designed according to an adequate model of dynamic behavior of aerodynamic surfaces th wing. The advantages of being able to implement the required levels of redundancy through the use of virtual sensors instead of some real sensors are obvious, in general, depending on the cost of physical sensors and problems in installing and exchanging data with the control system. These advantages are especially significant for sensors on board aerodynamic wing surfaces, in which the benefits are added in terms of weight and energy consumption. In this sense, experiments derived from the use of virtual sensors in the aviation field can be integrated, as described in "Rilevazione, isolamento e recupero of the guasti of the sensori di assetto di aeromobili", Graduation Thesis, Responsibles: M. Milanese (Dip. Automatica e Informatica), S. Chiesa (Dip. Ingegneria Aerospaziale), M. Birindelli (Alenia), Politecnico di Torino, 2003 by E. Corigliano, and in the automotive field, as described in "Experimental results in vehicle sideslip angle measurement" SAE 2006 , Detroit di M. Milanese, D. Regruto, e A. Fortina.

The numerical control unit 7b, which drives the aerodynamic surfaces of the wing, requires reliable and real-time acceleration and position information. In particular, three-dimensional accelerations that describe the behavior, in three-dimensional space, of the aerodynamic surface of a wing must necessarily be obtained on board the aerodynamic surface of the wing, i.e. on high.

This implies the need for the transmission system of the first portions of information 3a between the aerodynamic surface of the wing and the processing and control means 7 to meet strict specifications in terms of performance and energy absorption. In order to meet these requirements and, preferably, excluding the most obvious galvanic connection between the aerodynamic surface of the wing and the processing and control means 7 so as not to provide a favorable path for possible atmospheric discharges, the transmission system can be integrated into the drive cables of the aerodynamic surface of the wing by at least at least one data fiber.

Inserting optical fibers into cables, however, must take into account that drive cables typically have high modulus fiber and that the working environment is difficult for both Kevlar and UHMW polyethylene. As you know, Kevlar can leak and tend to absorb water, which can lead to increased electrical conductivity in case of acid rain or local pollution, leading to the need to use protective covers or braids, for example, made of THFT, which also perform the function of preventing abrasion. In this case, the natural placement of the optical fiber should be between the casing and the fiber bundle, paying attention to the provision of a certain freedom in length so as not to expose the optical component to the same elastic curvature of the cable.

In the case of UHMW-polyethylene, the considerations given for Kevlar remain valid, but the problem of its application should be added: the so-called "creep", i.e. irreversible elongation over time under the action of forces, which leads to frequent replacement of cables equipped with optical fibers, thereby reducing the actual working time. However, there are materials associated with polyethylene that reduce the problem, which can be considered in combination with the possible interlacing of the optical fiber inside the UHMW beam. Nevertheless, it must be emphasized that the cross-section increases with the square of the diameter, and therefore, the working load on the cables should be easily measured depending on the maximum required force, without subjecting to creep and without increasing the drag force, i.e. the resistance that the cables exert when air enters. Moreover, geometric modeling of the aerodynamic surface of the wing over the cross-section of the cable can be a useful solution for taking shape with less resistance and a large rise.

Other optical fibers in the cables can be used to power the first means of detection on board the aerodynamic surface of the wing. When introducing, in a low loss multimode fiber, a sufficient amount of light on the ground side, the light can be converted through a microphotoelectric module, for example made of GaAS, on the side of the aerodynamic surface of the wing.

Alternatively, the transmission system can provide the transmission of the first portions of information by radio frequency, and this solution, of course, is considered the most natural way of exchanging data, which requires the exclusion of electrical connections, but can be costly from an energy point of view.

Even if the optical fibers allow the transmission of information, in any case it is preferable to keep the additional resource represented by radio frequency transmission for backup purposes. If radio communication, therefore, provides a backup solution regarding the exchange of data via optical fibers, the security circuit may signal its activation, being attentive to possible interruptions in the acknowledgment of the receipt of optical streams.

Obviously, the radio frequency transmission system can use a huge variety of communication protocols in order to send the first portions of information to the ground processing and control means. When using, for example, the unidirectional continuous flow control protocol, the lower level of communication, the physical layer, there must be a radio frequency carrier, which can be a simple FSK modulator of digital information provided by sensors and always active, however, implying continuous energy absorption. Nevertheless, some solutions can be provided to reduce the necessary time and power for transmitting information, for example, the following:

- directional antennas with high gain: an antenna that does not add resistance to the air flow can be obtained using conclusions that correspond to existing geometry: suitable places can be a piece of cable near the aerodynamic surface of the wing or the wall of the aerodynamic surface of the wing. These two locations should have the advantage of being constantly directed towards the drive unit, which can accommodate an appropriate receive antenna that is communicating with the processing and control means;

- carrier suppression: this is a solution known as single sideband transmission, which provides greater energy savings, however, reducing the bit rate at which transmission can take place;

- asynchronous activation: this solution requires special software on board the transmitting device, which estimates the content of the data stream, transmitting information only at the most significant points in time due to the adoption of the concept of key frames of video compression. This advantage is important since energy reduction factors can be obtained that can be calculated similarly to data compression factors.

Alternatively, a packet datagram protocol can be used, for example, that used on the Internet to transmit data streams without characteristic sequence and validation, suitable for video and broadcasting. Since the protocol is bi-directional, the load from checking the rhythm of data exchange and the associated electrical absorption can be transferred to the control unit of the receiving station, which can only interrogate the sensors if necessary or rebuild the dynamic model of the control system according to the real state.

Alternatively, an asynchronous negotiation transport protocol can be used, which is more difficult to implement, but it can combine all the advantages of the above solutions. In essence, it is a very lightweight and susceptible bidirectional protocol that allows you to start exchanging data both from the side of the first detection tool, and from the processing and control tool. The lack of a stack leads to the fact that there are no delays that can reduce the bit rate.

In an additional alternative, the provision of the use of an ultrasonic transmission system is possible.

In any case, two accelerometers together with a magnetometer on board the aerodynamic surface of the wing form seven information flows at thousands of samples per second. This stream of raw data from the aerodynamic surface of the wing to the ground, in addition to being exorbitant for transmission, is practically useless for the geometric motor: in practice, the geometric motor must rotate at a frequency consistent with the size of the system and time constants, continuously providing updated position data in block of numerical control, and claims more reasonable data as input. For this purpose, the use of pre-processing means 11 may be provided, which is capable of pre-processing all or part of the first portions of information 3a on board the aerodynamic wing surface to provide pre-processed first portions of information 3a ', configured to transmit and simplified processing by means of a geometric motor 7a. For this purpose, accelerometers may be equipped with an integrated DSP (Digital Signal Processing) means 11.

In addition, as previously indicated, MEMS-type accelerometers used to collect useful information to obtain information about the flight path of the aerodynamic wing surface are sensitive to both static accelerations (gravity) and dynamic accelerations. Due to the need to use (static) acceleration of gravity to measure useful angles to obtain the position and orientation of the aerodynamic surface of the wing, the problem arises of isolating static acceleration from the intense aerodynamic accelerations to which the aerodynamic surface of the wing is subjected. This work can be done using the appropriate algorithm, which must be cyclic at speeds that are incompatible with the transmission speed available through the transmission system, and therefore, must be performed by pre-processing means 11 on board the aerodynamic surface of the wing.

The system 1 according to the present invention further comprises at least one power supply system of the first detection means and components of the transmission system on board the aerodynamic surface of the wing; it is obvious that this is the first tool and the components can be autonomously powered by their own batteries.

Nevertheless, it is necessary to take into account the fact that the system 1 according to the present invention, first of all, if used in connection with a "carousel" system, may require very high energy autonomy in order to eliminate the need to direct the aerodynamic surface of the wing to the ground with excessive frequency to replace or recharge the batteries, with the resulting overhead associated with the need to stop the carousel system. Moreover, it is useful to consider that system 1 should be able to work in conjunction with natural forces and events, such as rain, snow, ice, severe wind changes, atmospheric discharges, day, night, sun. To make available the limited energy requirement aboard the wing aerodynamic surface, alternative self-powered solutions are provided that utilize solar and visible wind elements. The power system, in fact, may contain plastic-based thin-film photovoltaic modules that can be advantageously used in the aerodynamic surface of a wing without modifying its aerodynamic characteristics and weight. These modules, obviously, should create enough energy to power the on-board electronics, increasing the ability to recharge potential batteries overnight.

Alternatively, you can use the explicit wind flow, which is always available around the aerodynamic surface of the wing; the power system, therefore, may contain at least one wind microturbine connected to a small generator with permanent magnets and fed by the wind.

The present invention further relates to a process for automatically controlling the flight of aerodynamic surfaces of a power wing, preferably by a system 1 similar to that described previously. In particular, the process according to the present invention operates in a predictive mode according to a “preferred control strategy” based on the observation and prediction of future flight conditions of the aerodynamic surfaces of the wing, taking into account critical situations and errors caused by forecasting, which allows avoiding local maxima, oscillations and instabilities of control of the aerodynamic surfaces wings. As described above, the system 1 according to the present invention is configured so that the processing and control means 7 receives as input information: position, accelerations, forces, and other geometrically defined quantitative quantities, processes them, and works as an output with winches that control the flight path aerodynamic surface of the wing.

In order to better describe the logic of the process according to the present invention, it is useful to refer again to the dynamic behavior of the aerodynamic surface of the wing. With reference, in particular, to FIG. 3, it is possible to schematically note the conditions of the wind front or wind flow W, which the wing aerodynamic surface 2 is capable of intercepting, at the following moments, relative to the ground, common with the limit point of the drive cables 21. Figure 3 describes, in fact, a quarter of the sphere, which is the navigation region of the aerodynamic surface 2 of the wing, in the center of which the so-called "force zone" 31 is defined, in which the aerodynamic surface 2 of the wing gives maximum thrust to the drive cables 21. When leaving " force zone "31 through the window 32, in which the aerodynamic surface 2 of the wing can move, in which the traction on the drive cables 21 is gradually reduced, the edge 22 of the window is reached, in which the aerodynamic surface 2 of the wing can move, while and on the drive cables 21 is significantly reduced.

With reference to FIGS. 4a and 4b, we now present, in this case, to consider the aerodynamic surface 2 of the wing placed in the center of the ideal “target” plane P, uniquely defined as normal to the bisector of the angle formed by the two drive cables 21. The process according to the present by working with the system 1, the invention can decide whether the wing aerodynamic surface 2 should perform any of the possible flight paths TV ₁ , TV ₂ , TV ₃ , ..., TV _n based on the current position in the center of the plane P. On this plane P can divide the flight path which can be followed by 2 aerodynamic surface of the wing, depending on the required time T _0, T _1, T _2, ..., T _n, to reach every point. In particular, the following time intervals can be considered, which correspond to the following angular positions of the lever 20a of the carousel system 20. The flight path of the wing aerodynamic surface 2 thereby occurs synchronously with the movement of the lever 20a.

For simplicity, we will consider, in particular, fig.4b with coordinates associated with the aerodynamic surface 2 of the wing. The Cartesian coordinate system is common with the aerodynamic surface 2 of the wing, and with it it moves in space. Therefore, the aerodynamic surface 2 of the wing is always in the center of the plane P. Only time flows. The diagram does not show the point at which it should come, as this is a future assessment. When the first time interval T ₀ expires, the coordinate system, common with the aerodynamic surface of the wing, again draws a “target”, which is completely similar to the previous one with the only difference that the time T ₁ became T ₀ , etc. T ₁ represents, therefore, a set of points that can be achieved by the aerodynamic surface 2 of the wing in its flight path in 1 step, T ₂ is a set of points that can be achieved in 2 steps, and the like. In any case, the aerodynamic surface 2 of the wing is always in the center of the “target” plane P.

For simplicity, FIGS. 2a and 2b show, by way of example, the “target” plane P up to time T ₂ , but it is obvious that the number n of steps that can be observed may be different.

In order to evaluate their own control strategies and select the flight path that the wing aerodynamic surface 2 should perform, the process according to the present invention uses flight and control parameters.

Therefore, with reference to FIG. 5, a “target” plane P can be noted on which some flight and control parameters according to the present invention are graphically included for the process.

Figure 5 shows how the parameters in a simplified form vary in space, reducing the complexity of the rotational movements of the "carousel" system in the model, common with the coordinate system of the aerodynamic surface 2 of the wing. The morphology and characteristics of these parameters are an essential part of the information that provides control for deciding on the flight strategy of the aerodynamic surface 2 of the wing. In fact, one can imagine the optimal movement, direction and position in this way to achieve the ideal height Q, in order to have the best wind, or with the best angle of attack relative to the wind, but at the same time, other parameters can be presented, such as, for example, the required maximum thrust area T, forbidden areas ZI (areas in which, for example, situations of unacceptable structural loads, instability, excess forces), and functional parameters of system 1, for example, areas in which the counterweight C enjoy thi s parameters in order to maintain the aerodynamic surface of the wing 2 and constant flight for rapid length adjustment of the drive ropes (dynamic length), supported on half of their own dynamics. Schedule C corresponding to the counterweight, for example, may allow suitable positions for lifting the counterweight or lowering it. In addition, the winch of the drive unit can obey the parameters, because they show the absolute length of the cables. The graph of the height parameter Q, on the contrary, represents the optimal area for height problems. The graph of the M maneuver parameter M, on the contrary, represents the optimal area for the most important maneuver in flight of the aerodynamic surface 2 of the wing, defined as an azimuthal turn through the fortewind, which consists in a sudden maneuver during which the aerodynamic surface 2 of the wing undergoes a rapid transition between the beam traverses. With specific reference to FIG. 6, it can be noted that in general, if the aerodynamic surface 2 of the power wing is free from scanning the wind front, not only the thrust by which it can drive the drive cables (and, as a result, the levers of the "carousel" system 20 ), far exceeds the thrust applied to keep the aerodynamic surface of the wing stationary at the point of maximum wind resistance (aerodynamic resistance), but also the area in which the aerodynamic surfaces 2 of the wing have a drag effect when rotated and "carousel" system 20 is completely excluded. In the area 37 against the wind or boulin, the ability to control the flight of the wing aerodynamic surface 2 allows you to perform an azimuthal rotation through the forwind, which consists in a quick transition between two traverses 36 and 38, during which the wing aerodynamic surface 2 passes at least a distance in the air three times the arc of a circle 34, spent in time, during which the "carousel" system 20 passes along this arc. Flight control should take into account that maneuver, in addition to being fast, does not have any negative impact on energy generation. 6, the position of each aerodynamic surface of the wing is random, i.e. the image should be considered as the top shot of the carousel system. In this configuration, each wing aerodynamic surface is free from searching for maximum wind intensity, avoiding the depleted areas of the wind front to pass immediately preceding the wing aerodynamic surface.

The size of each graph (weight, counterweight, etc.) is proportional to the allowed tolerance for the corresponding parameter. Each parameter, in turn, has a relative weight P _Q , P _C , P _M , P _ZI , P _T , relative height relative to all heights, which is described below.

From the plane shown in FIG. 5, after performing the maneuver and the expiration of the time interval T ₀ , we proceed to the new plane P, recalculated for the next solution. If global situations are unchanged, the charts related to each parameter tend to concentrate. Choosing a graphical representation of the optimal areas for each parameter considered is a way of resolving uncertainties and understanding which solution should be chosen. This strategy is useful in that it does not fall into local maxima, namely, positions that seem to be the best, but are not.

The process always has available information in direct form or in the form extracted from the first and second pieces of information detected using the detection system 1, associated with the flight altitude of the aerodynamic surface of the wing, with the dynamics of the counterweight, with traction values, with the calculation of safety in prohibited areas , with the time at which maneuvers should be performed. Let us consider, for example, the most important maneuver in flight of the aerodynamic surface of a wing, which is defined as an azimuthal turn through a fortewind. The decision on whether to execute it can be an event that is triggered: under such conditions, in fact, the process according to the present invention can provide an emergency step that demonstrates the aerodynamic surface of the wing, the maneuver of which should be carried out with the highest priority. It cannot be ruled out that a visionary strategy automatically thinks out a maneuver without the need for its proposal. If a good temporal depth of assessment is achieved, the turn through the fortwind is performed at the ideal time, calculated depending on the current information and parameters, since it is preferably the most reasonable action in order not to lose traction, not to lose dynamics on the counterweight and correspond to a specific height without hitting to prohibited areas.

The process step, which determines at each stage which direction the flight path of the aerodynamic wing surface can take, can be visually represented as a matrix, for example, shown in the following table. 1, containing for each parameter the best XY coordinates at times T ₀ , T ₁ , T ₂ , ..., T _n on a normal plane relative to the center of gravity of the drive cables.

Table 1 Parameter The weight T ₀ T ₁ T ₂ T _n Height P _Q QX ₀ Y ₀ QX ₁ Y ₁ QX ₂ Y ₂ QX _n Y _n Counterweight P _c Cx ₀ y ₀ CX ₁ Y ₁ CX ₂ Y ₂ Cx _n y _n Thrust P _T TX ₀ Y ₀ TX ₁ Y ₁ TX ₂ Y ₂ TX _n Y _n Prohibited area P _zi ZIX ₀ Y ₀ ZIX ₁ Y ₁ ZIX ₂ Y ₂ ZIX _n Y _n Maneuver P _M MX ₀ Y ₀ MX ₁ Y ₁ MX ₂ Y ₂ MX _n Y _n Resulting RX ₀ Y ₀ RX ₁ Y ₁ RX ₂ Y ₂ RX _n Y _n PT PT ₀ PT ₁ PT ₂ PT _n

The matrix in table 1, thus, contains the required data. The difference between the current height and the required height allows you to make one call to the logic of the block of numerical control or to calculate errors. This characteristic is characteristic of the matrix: in essence, there are current coordinates and the required coordinates for T ₀ , T ₁ , T ₂ , ..., T _n . All the parameters interpreted in the matrix create pairs of XY values for time moment T ₀ , T ₁ , T ₂ , ..., T _n . After that, the process photographs the situation in which the aerodynamic surface of the wing is located, and converts the plane P in Fig. 5 into numbers and coordinates. Given, for example, the height problem, the value of the matrix QX ₀ Y ₀ indicates a point next to the height parameter tending to the center of the graph of height Q in FIG. 5. The circular shape makes everything consistent in time when the trend is obvious: an increase in T ₁ , T ₂ . In the case of traction, the process can already calculate the development of the ideal point in time: the form of the required data is therefore not round.

Obviously, the relative weight P _Q , P _c , P _M , P _ZI , P _{T of} each corresponding parameter Q, C, M, ZI, T can be set, and this task can be dynamic (with feedback). By analyzing, for example, the average error of the required height relative to the current height, etc. the process can receive data on the most difficult parameters that must be satisfied. The feedback process adjusts the weights of the most critical parameters in such a way as to make decisions on these parameters more important. This type of error can be assigned an indicator, for example, a percentage indicator, standardized for each parameter relative to the maximum error value. For example, if in time the counterweight is always outside the center of dynamics and runs the risk of reaching its travel limit, this standardization provides the ability to detect which of the parameters makes the most errors. It can be an independent process that controls the weights of each parameter.

After collecting the best coordinates for each parameter, the process provides a stage at which the vector sum of all coordinates is calculated at time T ₀ . The resulting vector is RX ₀ Y ₀ , which is not yet the direction of the flight path along which the aerodynamic surface of the wing should move, since the forecast for the future should also be considered. The process then calculates the vector sums for all future stages RX ₁ Y _l , RX ₂ Y ₂ , ..., RX _n Y _{n n} , and then we introduce the time weights PT ₀ , PT ₁ , PT ₂ , ..., PT _n , which give priority to short-term strategies, while at the same time preventing the aerodynamic surface of the wing from entering potentially critical areas.

It is also obvious that the temporary weights PT ₀ , PT _l , PT ₂ , ..., PT _n can be set.

By calculating the matrix in Table 1, the process according to the present invention determines the ideal instantaneous coordinate (target) that should be sought when maneuvering the wing aerodynamic surface along the flight path. After finding the ideal coordinate, it is necessary to consider the maneuver and control of the drive cables in order to force the aerodynamic surface of the wing to achieve its goal. The process in this case includes the step of choosing the best path (the shortest path, with the passage of forbidden areas, etc.) in order to bring the aerodynamic surface of the wing from its current position to the target. Therefore, at this stage, the process selects, depending on the goal to be achieved, the best flight path to achieve it with minimizing time, since finding the aerodynamic surface of the wing is always correct on the target sequence defined as the best, and is a guarantee of generating maximum energy with maximum safety and maximum compliance with dynamic specifications. The essence of the problem at this stage is how to convert the coordinates of the target into traction. The step of choosing the best path, therefore, uses an inertial navigation system (INS) supported by the dynamic model of the aerodynamic surface of the wing (FVM), taking into account the equations and inertia of the flight of the aerodynamic surface of the wing, along with the percentage of the reaction that it may have depending on the difference in traction on ropes. Inertia and traction describe the law of maneuvering the aerodynamic surface of a wing; it is acceptable to take into account (predictive) estimates of the best path by evaluating all possible paths and evaluating maneuver using the decision tree. At this stage, apparent speeds and traction are taken into account, and you can accurately evaluate the best course strategy. The synergy between inertial navigation and the information provided through dynamic modeling, namely, motion models are obtained from the history of the position of the aerodynamic surface of the wing, from the control input data and from the forces acting on the aerodynamic surface of the wing itself, has been widely demonstrated previously using the dynamic equations of the vehicle used ( for example, with the help of Koifman and Bar-Itzhack, 1999; Ma et al., 2003). These studies demonstrate that the main advantage of using a vehicle model is to improve the ability to observe the sources of errors in INS.

In the inertial navigation system INS, the position (ρ _n ), speed (v _n ) and Euler angles (ψ) of the aerodynamic wing surface, referred to as the coordinate system n = [N, E, D] (north, east, down), are calculated as follows :

p _n = v _n

v _n = C _b ^m f _b + g _n

Ψ = E _b ⁿ ω _b

where g _n is the acceleration of gravity, f _b is the acceleration vector in three axes, ω _b is rotation. C ⁿ _b and E ⁿ _b are respectively the transformation and rotation matrices, defined as follows:

where s (.), c (.) and t (.) represent sin (.), cos (.) and tan (.), while Ψ = [φ, θ, ψ] are the Euler angles.

A dynamic model of the aerodynamic wing surface (FVM) with six degrees of freedom is instead composed of a set of equations that provides for the state variables of the aerodynamic wing surface, consisting of position, speed, Euler angles and rotations by means of control variables of the aerodynamic wing surface, which are accepted as known from system 1 The motion of the aerodynamic surface of the wing can be described by the following system of equations of motion, in which the forces acting on The vehicle is a function of the position, speed, Euler angles and current rotation of the aerodynamic surface of the wing:

where v _b = [u, v, w] are the velocity components along three axes in the coordinate system of the aerodynamic wing surface, ω _b = [p, q, r] are the rotations of the aerodynamic wing surface, F _x , F _y , F _z and I, M, N are the components of the force and moments acting on the aerodynamic surface of the wing along its own axes, g _x , g _y , g _z are the components of the acceleration vector of gravity expanded in the coordinate system of the aerodynamic surface of the wing, the mass of which is denoted like m. Coefficients C _0-9 are obtained based on the inertial matrix I.

For example, there can be two ways in which a dynamic model of the aerodynamic surface of the wing of the FVM can be applied as support to an inertial navigation system: the first method involves comparing and adjusting the speed and position of the aerodynamic surface of the wing, for example, obtained independently of the INS and the FVM. The second method uses the acceleration and rotation forecast performed by the FVM to implement direct accurate determination of the inertial measurement unit (IMU). In both cases, the INS processes the position, speed and Euler angles (which describe the rotation) of the aerodynamic surface of the wing to integrate the acceleration and rotation measurements provided by the IMU on board the aerodynamic surface of the wing. However, in the first method, the model of the aerodynamic surface of the wing calculates the speed and angles of the aerodynamic surface of the wing by using the control input of the aircraft itself. Moreover, the actual execution of FVM and INS takes advantage of the latest developments in quaternion-based mathematics. The task of the generalized Kalman filter (EKF) is to estimate the errors of the INS and FVM by observing the differences between the speed and angle data, respectively, generated by the INS and FVM.

In the second method, on the contrary, FVM is used to calculate the estimates of acceleration and rotation directly from the control input data. The input of the generalized Kalman filter is therefore made up of the differences between the acceleration and rotation estimates calculated by the FVM and the estimates read from the sensors used. The EKF is therefore used to evaluate the acceleration and rotation errors of the sensors and FVM, which are then used to subsequently adjust the sensors and FVM.

However, maneuvering the aerodynamic surface of a wing has a problem of accurate determination. Obviously, it is possible to choose the magnitude of the maneuver, but the magnitude must still be specified. In fact, there is a risk of oscillations with excessive amplification, due to internal reasons, the elasticity of the kinematic chain (winches are on the ground, maneuvers are performed in the air) and measurement delays (negligible). Consequently, there is a risk of inaccurately defined, insufficient or complicated maneuvers that make constant adjustments (opposite compensation) with the risk of uncontrolled fluctuations. To solve this group of problems, the field of control technology has already developed techniques, such as the Hinf methods and the Kalman filters already mentioned, which consider the delay of actuation as one of the disturbances, one of whose noise the control should regulate by optimizing the maneuver and its limitation using filters and methodologies that are calibrated in the system or self-calibrated. The described process according to the present invention can be equipped with prediction capabilities, the temporal depth of which is a function of the information processing power of the system according to the present invention. Another main characteristic that predicts the problems described above is that the processing and control means receives information related to acceleration. Excessive maneuver, therefore, is perceived for the most part in advance as to when the movement occurs, and this should lead the system into a subcritical situation where the oscillations may not be triggered by the ability of the sensors to provide 180 ° data before movement. These provisions should be directly accessible, and a new action should be performed after the damage has occurred, while acceleration notifies the damage.

Based on the above, the process according to the present invention, therefore, comprises the steps of:

a) detecting the first pieces of information 3a by means of the first detection means 3 related to the current moment of the flight path of the wing aerodynamic surface; possibly pre-processing all or part of the first pieces of information 3a by means of the pre-processing 11 to obtain pre-processed first pieces of information 3a ';

b) detecting second pieces of information 5a by means of second detection means 5a related to the current moment of the flight path of the wing aerodynamic surface;

c) sending by means of a transmission system of the first pieces of information 3a, 3a 'to the processing and control means 7, in particular to the geometric motor 7a;

d) sending second pieces of information 5a to the processing and control means 7, in particular to the geometric motor 7a;

e) from the first 3a, 3a 'and second pieces of information, directly or indirectly obtaining values related to at least the current XY position and current flight altitude of the wing aerodynamic surface, the counterweight dynamics and the traction in the drive cables;

f) determining flight and control parameters, such as, for example, altitude Q, counterweight dynamics C, maneuver M, forbidden areas ZI, thrust T of power cables; possibly determining tolerances for each of these parameters;

g) determining the relative weight P _Q , P _C , P _M , P _ZI , P _T for each of the flight and control parameters;

h) calculating for each parameter the best XY coordinates at the following time instants T ₀ , T ₁ , T ₂ ,. .., T _n ;

i) calculating the vector sum RX ₀ Y _{0 of} all coordinates at time T ₀ ;

j) calculating vector sums RX ₁ Y ₁ , RX ₂ Y ₂ , ..., RX _n Y _n for all future time instants T _l , T ₂ , ..., T _n ;

k) the definition and application of temporary weights PT ₀ , PT ₁ , PT ₂ , ..., PT _n for vector sums;

l) choosing the best of RX ₁ Y ₁ , RX ₂ Y ₂ , ..., RX _n Y _n as the ideal instantaneous coordinate (target) to which the maneuver of the wing aerodynamic surface should aim;

m) choosing the best path for the flight path of TV ₁ , TV ₂ , TV ₃ , ..., TV _n so as to direct the aerodynamic surface of the wing from its current position to the target;

n) directing the aerodynamic surface of the wing from the coordinates of the current position to the target by actuating the numerical control unit 7b of the system 1; preferably using a navigation system (INS) supported by a dynamic model of the aerodynamic surface of the wing (FVM);

o) repeating steps a) to n) for each time interval Δt, defined as the frequency of the control loop; it can be ensured that the process according to the present invention also comprises a feedback adjustment step of duration Δt, making the frequency of the loop de facto adjustable. If Δt is short, the process performs an accurate and detailed, but possibly redundant and hardly far-sighted short-term forecast analysis, since n possible time steps do not reach the globally optimal time interval from the current moment: it becomes necessary to link the stages of future prediction with the duration of the step in order to optimize the temporal depth of the flight path of the aerodynamic wing surface to be performed. Therefore, it is necessary to evaluate the possibility of fulfilling a forecast that covers a too distant future: it is reasonable to assume that providing more than one turnaround of a "carousel" system is useless, since the situation repeats itself in the conditions of stability. The ideal duration Δt is possibly a duration that corresponds to the duration of a complex maneuver, such as azimuthal rotation through the fortewind of the aerodynamic surface of the wing.

Claims (36)

1. A system (1) for automatically controlling the flight of at least one aerodynamic surface (2) of a power wing, said aerodynamic surface (2) of a power wing being controlled by a drive unit (9) equipped with two winches to which said aerodynamic power wing surface (2) by means of two corresponding power cables (21), characterized in that it comprises: first means (3) for detecting on board said power wing aerodynamic surface (2), made the ability to detect the first pieces of information (3a) related to at least one position and one orientation in the space of the said aerodynamic surface (2) of the power wing and the accelerations to which the mentioned aerodynamic surface (2) of the power wing is subjected, second means (5) detection on the ground, configured to detect second pieces of information (5a) related to at least one tension value of said drive cables (21) of said aerodynamic surface (2) of forces wing and to the position of the counterweight of said drive unit (9), means (7) for processing and controlling said first (3a) and said second pieces of information (5a), configured to convert the contents of said pieces of information (3a, 5a) into a mechanical drive , operating said winch the said drive unit (9) for driving said aerodynamic surface (2) power wing path TV _1, TV _2, TV _3, ..., TV _n flight maximizing the effect of "lift" generated by said AER dynamic surface (2) power wing by wind stream W, in which it is immersed, and maximizing the amount of kinetic energy subtracted in said wind stream W, and a transmission system of said first pieces of information (3a) in said means (7) processing and management. 2. System (1) according to claim 1, characterized in that said processing and control means (7) are arranged to process said first pieces of information (3a) in order to return information (7 s) about position, acceleration and orientation said aerodynamic surface (2) of the power wing into a numerical control unit (7b) configured to operate (9a) with said winches of said drive unit (9) to control the traction force on said drive cables (21). 3. System (1) according to claim 2, characterized in that said transmission system transmits said first portions of information (3a) to said processing and control means (7). 4. System (1) according to claim 1, characterized in that said first detection means (3) comprises triaxial accelerometers. 5. System (1) according to claim 4, characterized in that the said triaxial accelerometers are devices such as microelectromechanical systems (MEMS). 6. System (1) according to claim 1, characterized in that said first detection means (3) comprises an electronic compass. 7. System (1) according to claim 6, characterized in that said electronic compass is a flux-gate magnetometer. 8. System (1) according to claim 4, characterized in that each one of the two mentioned three-axis accelerometers is located on the corresponding end of the said aerodynamic surface (2) of the power wing, next to the connection of the said drive cables (21) with the walls of the said aerodynamic surface ( 2) power wing. 9. The system (1) according to claim 1, characterized in that said second detection means (5) comprises strain gauges configured to measure deviations of said drive cables (21). 10. System (1) according to claim 1, characterized in that said second detection means (5) comprises encoders on said winches of said drive unit (9). 11. The system (1) according to claim 1, characterized in that said second detection means (5) comprises proximity sensors. 12. System (1) according to claim 1, characterized in that said second detection means (5) comprises a ground-based artificial vision system. 13. System (1) according to claim 1, characterized in that said aerodynamic surface (2) of the power wing is made of rheological polymers. 14. System (1) according to claim 1, characterized in that said first (3) and / or second (5) detection means comprise virtual sensors. 15. System (1) according to claim 1, characterized in that said transmission system is integrated into said drive cables (21) of said aerodynamic surface (2) of a power wing via at least one data transmission optical fiber. 16. System (1) according to claim 1 or 15, characterized in that the said drive cables (21) have a cross section of the shape of the aerodynamic surface (2) of the power wing. 17. System (1) according to claim 1, characterized in that said transmission system is of the radio frequency type. 18. System (1) according to claim 1, characterized in that said transmission system is of the ultrasonic type. 19. System (1) according to claim 1, characterized in that said transmission system uses a unidirectional continuous stream control protocol. 20. System (1) according to claim 1, characterized in that said transmission system uses a datagram packet transmission protocol. 21. System (1) according to claim 1, characterized in that said transmission system uses asynchronous negotiation transport protocol. 22. System (1) according to claim 1, characterized in that it comprises a pre-processing means (11) configured to pre-process all or part of the first pieces of information (3a) on board the aerodynamic surface (2) of the power wing, to provide pre-processed first pieces of information (3a '). 23. The system (1) according to claim 1, characterized in that it comprises at least one power supply system of said first detection means (3) and said transmission system on board said aerodynamic surface (2) of the power wing. 24. System (1) according to claim 23, characterized in that said power system comprises plastic-based photovoltaic thin-film modules superimposed on said aerodynamic surface (2) of a power wing. 25. System (1) according to claim 23, characterized in that said power system comprises one wind microturbine connected to a generator with permanent magnets. 26. The process of automatic flight control of at least one aerodynamic surface (2) of the power wing by means of system (1) according to any one of the preceding paragraphs, characterized in that it comprises the steps in which: a) the first portions of information are detected (3a) ) by means of the aforementioned first detection means (3) related to the current moment of the flight path of said aerodynamic surface (2) of the power wing, b) detecting said second pieces of information (5a) by means of said second The detection means (5a) related to the current moment of the flight path of the aforementioned aerodynamic surface (2) of the power wing, c) send the said first pieces of information (3a, 3a ') to the said processing and control means (7) through the mentioned transmission system, d) send said second pieces of information (5a) to said processing and control means (7), e) from said first (3a, 3a ') and second pieces of information, directly or indirectly receive values related to at least one current position XY and one current altitude flight of said aerodynamic surface (2) of the power of the wing, to the dynamics of said counterweight and to thrust on said actuating wire (21), f) define the parameters of flight and control, g) determining the relative weight P _Q, P _C, P _M , P _Z1 , P _T for each one of the mentioned flight and control parameters, h) calculate for each one of the mentioned parameters the best XY coordinates at the following time instants T ₀ , T ₁ , T ₂ , ..., T _n , i) calculate vector sum RX ₀ Y ₀ of all coordinates at said time T _0, j) Calcd r vector sums RX ₁ Y _1, RX ₂ Y _2, ..., RX _n Y _n for all future time points T _1, T _2, ..., T _n, k) is determined and temporal weight RT ₀ is used, PT _1, PT ₂ , ..., PT _n for the said vector sums, 1) choose the best one of the said vector sums RX ₁ Y ₁ , RX ₂ Y ₂ , ..., RX _n Y _n as the ideal instantaneous coordinate (target) that the aerodynamic maneuver tends to surfaces (2) of the power wing, m) choose the best path of the flight paths TV ₁ , TV ₂ , TV ₃ , ..., TV _n to direct the aforementioned aerodynamic surface (2) of the power wing from the current position I to the mentioned target, n) direct the mentioned aerodynamic surface (2) of the power wing from the mentioned coordinate of the said current position to the said target, acting on the said drive unit (9) by the said block of numerical control (7b), o) repeat steps a) -n) in each time interval Δt. 27. The process according to p. 26, characterized in that it contains between step a) and step b) a stage in which all or part of the aforementioned first pieces of information (3a) is preliminarily processed by said pre-processing means (11) to obtain pre-processed first pieces of information (3a '). 28. The process according to p. 26, characterized in that the mentioned flight and control parameters are height Q, counterweight dynamics C, maneuver M, forbidden areas ZI, thrust T of said drive cables (21). 29. The process of claim 26, wherein said step f) comprises a step in which a tolerance is determined for each one of said parameters. 30. The process of claim 26, wherein said step m) uses an inertial navigation system (INS) supported by a dynamic model of said wing aerodynamic surface (2) (FVM). 31. The process according to p. 26, characterized in that it contains an emergency step to indicate the aerodynamic surface (2) of the power wing, which maneuver should be performed with maximum priority. 32. The process according to p. 26, characterized in that it contains a stage in which the above-mentioned parameters of flight and control are adjusted with feedback. 33. The process according to p. 26, characterized in that it comprises the step of precisely determining said maneuver by means of Hinf techniques and / or Kalman filters. 34. The process according to p. 26, characterized in that it contains a stage in which the duration of the mentioned time interval Δt is regulated with feedback. 35. The use of system (1) according to any one of claims 1 to 25 in conjunction with a "carousel" system (20). 36. Application of the process according to any one of paragraphs.26-34 in conjunction with a “carousel” system (20).

Images