This document relates to the feedback control of an optronic sight for a motorized vehicle such as a propeller-driven aircraft, a propeller-driven boat or a tracked vehicle.
An optronic sight 2 consists of a set of cameras and/or pointing devices, called aiming module 4. This aiming module 4 is placed on a support 6 of a motorized vehicle and can move according to two axes 8a, 10. The line of sight 12 of said optronic sight 2 refers to the optical axis emerging from one of these sensors. The purpose of the optronic sight 2 is to orient the line of sight 12 towards a target regardless of the movements of the motorized vehicle and/or of the target, and regardless of the external environment (atmospheric conditions, etc.). To this end, said aiming module 4 includes means 14 for continuously measuring angular data, i.e. a gyrometer 14 in the case of measurement of the angular speed or a gyroscope 14 for measuring the angular position of the line of sight 12 as illustrated in
The rotational speed of the rotor and the blades of a helicopter or the engine speed of propeller-driven planes, propeller-driven boats or tracked vehicles generate vibratory disturbances which deteriorate the stabilization of the line of sight 12 of the optronic sights 2. It is then necessary to set up a process allowing accurately stabilizing the image and therefore in particular to correct the angular data (angular speed or position) of the line of sight 12 thanks to a corrector 15. Afterwards, this correction is made by means of control means 16 of the moving means 17a, 17b which may comprise gimbals actuated by motors. To reject the vibratory disturbances acting on the aiming module 4 and thus make the line of sight 12 fixed in an inertial frame, it is therefore necessary that the sum of the torques, i.e. the motor torque Cmot and the friction torque Cf due to the bearings of the gimbals, applied to the aiming module 4 is zero.
For this purpose, it is conventionally known to use a feedback control loop 20 capable of acting on the angular data (speed or position) of the line of sight 12 as illustrated in
Thus, the feedback control function of the line of sight of the optronic sight 2 will be developed in more detail. The latter is split into two portions: an analog portion 22 and a digital portion 24. First of all, in the analog portion 22, the spectral lines associated with the disturbing vibrations γvib, generated by the rotation of the rotor and the blades of a helicopter are identified and filters set on these spectral lines will then be constituted. Thus, the transfer function Hvib allows modeling the impact of the disturbing vibrations γvib on the angular orientation of the line of sight. Hence, at the output of the transfer function, the angular disturbance δy of the line of sight due to the disturbing vibrations γvib is obtained which can thus be considered in the feedback control loop of the optronic sight. Hence, this control method is based on a priori knowledge of a model of the studied system.
Afterwards, another step consists in modeling the dynamics of the angular data (angular position or speed) of the line of sight by a transfer function called Hgyro. This transfer function is based either on the measurement y of the position of the line of sight by means of a gyroscope, or on the measurement y of the angular speed of the line of sight obtained by a gyrometer or more specifically by the inertial sensor of the gyrometer. Henceforth, the measurement of the angular data of the obtained line of sight ym then passes through an Analog-to-Digital Converter (ADC) and is thus sampled to become the sampled measurement ymk. A feedback control error εk is then obtained by the difference between a reference yck and the sampled measurement ymk. This feedback control error εk is then input into a linear and time-invariant corrector K. The latter is calculated in order to compensate for the disturbing vibrations γvib whose fundamental frequency fv is fixed over time. The software implementation of said corrector K is done in the form of a combination (sum and/or product) of second-order digital linear filters. At the output of this corrector is obtained a digital motor command uk which is afterwards transformed into an analog command u (voltage) by a Digital-to-Analog Converter (DAC). This analog command u is applied to the electric motor, modeled by the transfer function Hmot, which consequently delivers an electromechanical torque. Thus, it allows obtaining the electromechanical torque Cmot to be supplied by the motor to turn the gimbals. The greater the error εk, the higher the torque Cmot supplied by the motors will have to be in order to reduce this error. The electromechanical torque Cmot supplied by the motor actuates the gimbals modeled by the transfer function Hcardan, in order to compensate for/cancel the error εk. This error is due on the one hand to the disturbing torque of frictions in the bearings of the gimbals and on the other hand to the angular disturbance δy.
As regards aerial vehicles or propeller-driven aircrafts, propeller-driven boats or tracked vehicles, there is currently no effective solution for such motorized vehicles including, in operation, variations in engine speed or variations in the vibration frequencies generated by the tracks. Indeed, variations in engine speed induce a spectrum of vibrations with specific spectral lines whose fundamental frequency varies with said engine speed and which requires the use of correctors adapting to these variable frequencies.
In other words, there is no corrector allowing compensating for the angular movements of the line of sight due to the linear vibrations of the aircraft carrying the sight, when these vibrations have a spectrum with lines at variable frequencies. These angular movements may result from a deformation of the mechanics which are not infinitely rigid. It consists of a point of view of the line of sight with slight rotations at high frequency which induce a blur on the image obtained by a camera, even though in average value the line of sight actually points in the same direction. Hence, it is necessary to determine a feedback control loop guaranteeing some stability over an a priori predefined domain of variable frequencies associated with the vibrations of said aircraft. This determination should be made in real-time, i.e. in operation.
The present document aims to remedy the aforementioned drawbacks, in a simple, reliable and inexpensive manner.
To this end, the invention relates to an optronic sight for a motorized vehicle such as an aerial or marine vehicle propelled by a propeller or a tracked land vehicle comprising an aiming module able to be moved about a first axis and a second axis not parallel to the first axis, means for moving the aiming module about the first and second axes, means for continuously measuring an angular data of said module about the first and second axes, said optronic sight being characterized in that it comprises a feedback control loop comprising: means for continuously measuring the acceleration of the aiming module according to three orthogonal directions of the space, means for detecting at least one fundamental frequency of the vibratory disturbances generated by the operation of the motorized vehicle, this frequency being obtained based on the output data of said acceleration measuring means, an adaptive corrector configured to continuously receive as input said fundamental frequency, a discrepancy between an angular setpoint value and said angular data, and to output a movement setpoint value to the moving means.
By motorized machine, it should be herein understood that it consists of a vehicle generating in operation a spectrum of vibrations containing lines whose frequencies could vary during use.
Thus, the adaptive corrector varies according to the fundamental frequency of the spectrum of the lines of the disturbing vibrations, while guaranteeing the stability of the feedback control loop. This adaptive corrector allows automatically adapting in real-time the frequencies of the selector filter to be used to eliminate the vibratory disturbances due to the operation of the motorized vehicle. The vibrations correspond to slight rotations at high frequencies which induce a blur on the image obtained by a camera, even though in average value the line of sight points in the same direction. A more robust inertial stabilization of the line of sight, the maximum of performance and good robustness margins of the warning loop are therefore obtained. Furthermore, thanks to its design, said adaptive corrector can be used on several motorized vehicles. Only the frequencies specific to the motorized vehicle should be provided to the adaptive corrector, without any additional setting to be performed. In practice, this represents a significant time saving.
The means for continuously measuring said angular data may include a gyroscope able to obtain an angular position or a gyrometer able to obtain an angular speed.
Said means for measuring the acceleration may comprise an accelerometer.
The fundamental frequency may be obtained by Fast Fourier Transform followed by a calculation of a maximum in the frequency band obtained after carrying out said Fast Fourier Transform or by phase-locked loop.
The adaptive corrector may be a Linear Variant Parameter corrector.
This Linear Parameter Variant corrector is linear but varies over time, according to measurable parameters. It linearly depends on the variant parameter.
Said adaptive corrector may follow the state representation according to the following formula:
where xk is the state variable of the corrector, εk is the feedback control error at the input of the corrector, uk is the digital motor command calculated by the corrector (output of the corrector), fmin, and fmax are two frequencies limiting the fundamental frequency in real-time {circumflex over (f)}vk of the disturbing vibrations γvib.
The Linear Variant Parameter (LPV) corrector may comprise the following affine state matrices:
A(vk)=A0+vkA1
B(vk)=B0+vkB1
C(vk)=C0+vkC1
D(vk)=D0+vkD1
where A0, B0, C0, D0, A1, B1, C1, D1 designate matrix gains which are the parameters saved in the memory of software that implements said corrector.
The first axis and the second axis may be perpendicular to each other.
The present document also relates to a motorized vehicle such as a helicopter, an aerial or marine vehicle propelled by a propeller or a tracked land vehicle comprising an optronic sight as described hereinabove.
Similarly to the prior art, said optronic sight 2 comprises in particular:
In the embodiment illustrated in the figures, the first axis 8a and the second axis 10 are perpendicular but it should be understood that the details of embodiments given hereinafter are also applicable to embodiments wherein the axes are not perpendicular and not even secant. The first axis 8a and the second axis 10 may also be secant and not perpendicular. Said optronic sight 2 of
Phase-locked loops conventionally consist of a phase comparator, a loop filter, a voltage-controlled oscillator and a possible frequency divider.
As regards the calculations operated by the adaptive corrector K(vk), three techniques can be used: either by using a Linear Parameter Variant (LPV) control, or by means of a symbol corrector, or by a combination of its two types of correctors (LPV and symbol).
In the case of an LPV control corrector, a minimum state representation of the system K(vk) is designated by (A, B, C, D) with Aϵn×n, Bϵn×1, Cϵ1×n and Dϵ. . . The software implementation in the form of a state of the adaptive corrector K(vk) is done according to the following relationship:
where xkϵn is the state variable of the adaptive corrector, εk is the feedback control error at the input of the adaptive corrector, uk is the digital motor control calculated by the adaptive corrector (output of the adaptive corrector), fmin and fmax are two frequencies limiting the fundamental frequency in real-time vk of the disturbing vibrations γvib. The state matrices (A, B, C, D) are vk affine and are written in the form:
A(vk)=A0+vkA1
B(vk)=B0+vkB1
C(vk)=C0+vkC1
D(vk)=D0+vkD1
where A0, B0, C0, D0, A1, B1, C1, D1 designate matrix gains which are the parameters saved in the memory of software that implements said adaptive corrector K(vk).
Thus, the adaptive corrector K(vk) varies according to the fundamental frequency of the spectrum of the lines of the disturbing vibrations γvib (fv), while guaranteeing the stability of the feedback control loop. This adaptive corrector K(vk) allows automatically adapting in real-time (without having to recalculate the entire corrector unlike the prior art) the frequency of the selector filter to be used without any risk of instability, saturation or degradation of the robustness of the feedback control.
Number | Date | Country | Kind |
---|---|---|---|
2009793 | Sep 2020 | FR | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/FR2021/051652 | 9/24/2021 | WO |