Method for the automatic preparation and configuration of control laws for a remotely piloted multi-rotor aircraft

Abstract

A method for preparing a remotely piloted multirotor aircraft for a mission, the method comprising the steps of: a) obtaining at least one mission parameter and one optimization criterion, b) selecting a combination of aircraft components, c) obtaining characteristics of each component, d) calculating aircraft control law coefficients for the component combination by minimizing a function representing a trajectory tracking error, e) calculating a cost function value for the combination selected from the optimization criterion, f) applying a validation criterion to the calculated cost function value and, if the validation criterion is not met, selecting a new combination to minimize the cost function value and iterating steps c) to e), g) obtaining an optimal combination of components and control law coefficients. Figure to be published with the abstract: 4

Description

Title of the invention: Method for preparing and automatically configuring control laws for a remote-controlled multi-rotor aircraft Technical field of the invention

[0001] The invention relates to the fields of design and configuration of a remotely piloted multi-rotor aircraft, for example a vertical take-off and landing drone, for example a propeller drone, in particular a four-engine drone.

[0002] More particularly, the invention relates to a method for preparing a remote-controlled multi-rotor aircraft for a predetermined mission. State of the prior art

[0003] The piloting algorithms of remotely piloted multi-rotor aircraft are implemented by on-board control electronics and allow regulation of the altitude, position and attitude of the aircraft according to commands received remotely and / or recorded autonomous missions.

[0004] These algorithms operate by implementing one or more regulation loops for these quantities, which involve a plurality of regulation coefficients.

[0005] Today the process of adjusting these flight control laws for a given application is a long process specific to the configuration and specifications of the aircraft.

[0006] The adjustment and verification campaigns of these control laws include test cycles such as flight tests, physical simulations and tests of the electronics during the execution of the control algorithms.

[0007] The process of adjusting the control laws must be repeated when one or more of the components of the aircraft are removed or modified and / or when a new component is added. Several combinations of components having different characteristics from one component to another are generally conceivable for the same type of mission. In this case, adapting the aircraft to new operational conditions may require a significant adjustment time, by seeking the optimal combination of components for the aircraft and the associated control law adjustments, by successive tests. Presentation of the invention

[0008] The invention aims to overcome these drawbacks, by allowing rapid and efficient selection of the optimal components for a remote-controlled multi-rotor aircraft for a given mission, and adjustment of the control laws corresponding to this combination of components.

[0009] The invention aims in particular to enable automatic calculation of control gains for a new assembly of drone components. Preferably, the calculation of control gains can be carried out within the drone itself, in an automated and offline manner, thanks to automatic recognition within the drone of the components installed at a given time without necessarily communicating with an external server via a telecommunications network.

[0010] The invention thus falls within the objective of having a modular drone, capable of adapting to different operational requirements (performance criteria, mission profile, availability of components, etc.) in a simple and automated manner.

[0011] To this end, the invention relates to a method for preparing a remotely piloted multi-rotor aircraft for a mission, the method comprising steps of:

[0012] a) obtaining at least one mission parameter and one optimization criterion,

[0013] b) selecting a combination of components for the aircraft,

[0014] c) obtaining characteristics of each component,

[0015] d) calculation of coefficients of aircraft control laws for the combination of components, by minimizing a function representative of a trajectory tracking error, from each mission parameter and the characteristics of the components,

[0016] e) calculating a value of a cost function for the combination, said cost function being selected from the optimization criterion, from characteristics of the components, from each mission parameter and from the coefficients of the control laws,

[0017] f) applying a validation criterion to the value of the calculated cost function, and, if the validation criterion is not satisfied, selecting a new combination by modifying at least one of the components so as to minimize the value of the cost function, and iterating steps c) to e),

[0018] g) obtaining an optimal combination of aircraft components for the mission and coefficients of the control laws for the optimal combination, and

[0019] h) installation on the aircraft of the components according to the optimal combination and transmission of the coefficients of the control laws to an aircraft control electronics.

[0020] Such a method makes it possible to adapt an aircraft to new operational requirements quickly and efficiently by determining an optimal combination of components and control law coefficients.

[0021] Said components may comprise, for example, at least one structure, at least one motor, at least one rotor and at least one battery.

[0022] The validation criterion is for example a threshold value of the cost function, which represents a maximum cost to have a satisfactory combination.

[0023] Alternatively, the validation criterion may be a function of the value of the cost function and / or a derivative of the cost function.

[0024] The method may also comprise a step a') of obtaining general parameters relating to the aircraft and independent of the mission, said general parameters being taken into account in the choice of components to minimize the values ​​of the cost function, said general parameters preferably comprising a flight envelope outside the nominal mission.

[0025] Such a characteristic makes it possible to take into account usual constraints that are not specific to the mission in order to eliminate solutions that do not respect them.

[0026] Step c) of obtaining characteristics may comprise reading an identification element of each component and querying a database comprising characteristics of available components.

[0027] Such a feature allows a direct association between each component of the aircraft and its pre-recorded characteristics, in order to access them quickly and without risk of error.

[0028] The identification element may be a barcode or, preferably, a QR code.

[0029] Such a feature allows rapid reading of the identification elements by means of a suitable device, without risk of reading error.

[0030] Step c) of obtaining the characteristics may comprise obtaining at least one of a mass of the component and one or more specific functional characteristics of the component, such as an arm length of a structure, a pitch angle, a lift coefficient, a drag coefficient and / or a rotor blade length, nominal and maximum torque values ​​provided by motors, an inertia of a motor shaft and an electrical resistance of the motors, or a number of cells and / or a storage capacity of a battery.

[0031] Such a characteristic makes it possible to obtain real values ​​of characteristics of the different components which come closest to the optimal values ​​determined by the optimization algorithm and to use these real values ​​in the calculation of the control law coefficients and in the rest of the preparation process.

[0032] The cost function may be a function of a total mass of the aircraft and / or a function of a total energy expended by a propulsion system of the aircraft during the mission, and / or a function of a maximum power consumed by the propulsion system of the aircraft during the mission.

[0033] Such a feature makes it possible to maximize an aircraft payload or an aircraft endurance for long-duration flights.

[0034] The propulsion system includes, for example, the engines of the aircraft.

[0035] The cost function can be expressed as a linear combination of the mass total of the aircraft with a first coefficient, of the total energy expended by the propulsion system with a second coefficient and the maximum power consumed with a third coefficient.

[0036] This characteristic makes it possible to take into account the three optimization criteria in the cost function by assigning each of them a relative weight according to its importance for the type of mission planned.

[0037] Step d) of calculating coefficients of the control laws may include taking into account at least one external constraint linked to external conditions.

[0038] Such a step makes it possible to strengthen the resilience of the aircraft equipped with the set of coefficients determined by external conditions such as wind or gusts of wind.

[0039] During step e), the value of the cost function can be calculated from a simulation of a flight of the aircraft with the control law coefficients obtained in step d).

[0040] Such a characteristic makes it possible to reproduce the entire planned flight of the aircraft and thus to have a cost function representative of reality.

[0041] During step f), the minimization of the cost function can take into account at least one constraint linked to characteristics of the components.

[0042] Such a characteristic makes it possible to avoid solutions which would not be feasible in practice due to limiting characteristics of the components.

[0043] The invention also relates to a method for carrying out a mission by an aircraft, comprising the implementation of a method for preparing the aircraft as described above and a step of flying the aircraft with the components according to the determined optimal combination and provided with the associated control law coefficients. Brief description of the figures

[0044] [Fig. 1] is a schematic representation of a remotely piloted multi-rotor aircraft,

[0045] [Fig.2] is a graphical representation of a pre-recorded trajectory of the aircraft of [Fig.l],

[0046] [Fig.3] is a graphical representation of mission parameters corresponding to the trajectory of [Fig.2], and

[0047] [Fig.4] is a schematic representation of an aircraft preparation method according to the invention. Detailed description of the invention

[0048] An aircraft 1 is shown in [Fig.l]. The aircraft 1 is a remotely piloted quad-rotor aircraft, also known as a quad-engine drone.

[0049] The aircraft 1 comprises a structure 3 formed of a central body and four arms 6 distributed around the central body on either side, in a median plane P. The aircraft 1 also comprises four engines 5 arranged at the ends of the arms 6, each engine 5 being arranged to drive a rotor 7 in rotation around an axis of rotation Respective R perpendicular to the median plane P.

[0050] A roll axis X, a pitch axis Y and a yaw axis Z are defined, forming an orthogonal reference frame as shown in [Fig.l].

[0051] The motors 5 are electric motors controlled independently by means of an electric current, preferably determined to reach a rotation speed setpoint (speed control of the motors), and apply a mechanical torque to the rotor axis which results in mechanical thrust.

[0052] The aircraft further comprises at least one power supply device such as a battery 9 capable of supplying the motors 5 with energy.

[0053] The structure 3, the motors 5, the rotors 7 and the battery 9 are collectively referred to as components. Several models of these components are available and have varied characteristics. The arrangements of different models of components are hereinafter referred to as combinations.

[0054] Advantageously, each of the components is provided with an identification element, such as a barcode, a QR code or an RFID chip, for example, allowing an operator to simply identify the component and to make the link with a database comprising the characteristics of the component to obtain these characteristics directly and with a reduced risk of error.

[0055] The aircraft 1 further comprises control electronics comprising a computer 15, configured to determine the coefficients of the flight control laws of the aircraft 1 as a function of the characteristics of the components of the aircraft and a flight controller 17 configured to implement the control laws in order to determine the speed instructions transmitted to the engines 5 to comply with a predetermined trajectory and flight conditions.

[0056] According to a first embodiment, the computer 15 is on board the aircraft, and is adapted to calculate the coefficients of the control laws independently, without communication with an external device.

[0057] Alternatively, the computer 15 is partially remote, that is to say that it is adapted to communicate with at least one external computer to perform the calculation of the control law coefficients.

[0058] The computer 15 comprises, in a manner known per se, at least one processor and one memory.

[0059] The flight controller 17 also comprises at least one processor and one memory, which can be shared with the computer 15 or separate, as well as remote communication means configured to allow remote piloting.

[0060] The memory of the flight controller 17 is configured to store the values ​​of the control loop coefficients determined by the computer 15 for their implementation in the piloting of the aircraft 1.

[0061] The memory of the flight controller 17 also contains at least one pre-recorded autonomous mission. Each autonomous mission comprises a plurality of mission parameters, including in particular at least one pre-recorded autonomous trajectory, which can be expressed in terms of positions, speed, and / or translational acceleration for the aircraft 1. The pre-recorded trajectory can, alternatively or in addition to the preceding characteristics, comprise a plurality of waypoints of the aircraft during its journey. Each such waypoint is for example defined by three-dimensional coordinates, such as a position determined by GPS and an altitude, as well as by an absolute value of ground speed.

[0062] Compliance with these pre-recorded trajectories is ensured by the implementation of control loops for the associated quantities, in particular the position, speed, orientation and rotation speed, by the flight controller 17. These control loops involve known control laws, of the proportional corrector (P), proportional-integral corrector (PI) and proportional-integral-derivative corrector (PID) type, applied to correct a deviation from a setpoint value. Each of these correctors is governed by one or more gain coefficients associated with the proportional, integral and derivative terms of the deviation from the setpoint. These coefficients can be adjusted to best follow the regulation instructions, i.e. the trajectory in the present case.

[0063] For example, the position and orientation of the aircraft 1 may be governed by proportional corrector type loops, while the speed and angular rotational speed of the aircraft 1 may be governed by proportional-integral corrector type loops.

[0064] A trajectory is for example characterized by a climb speed, a flight altitude, a cruising speed and duration, and finally a descent speed.

[0065] An example of a trajectory is shown in [Fig. 2], which is a graphical representation of the altitude z of the aircraft 1 as a function of a ground distance traveled x. The curve T represents the desired pre-recorded trajectory, and the curve Tr represents the actual trajectory followed by the aircraft 1, obtained by measurement during a test or by simulation, with deviations corrected by the control loops.

[0066] Indicative examples of mission parameters are given in [Fig. 3], which includes variation curves as a function of time t, respectively of a linear ground speed Vx, a linear ground position x, a transverse ground speed Vy, a transverse ground position y, a vertical speed Vz and a vertical position z.

[0067] An example of a method for preparing the aircraft 1 for a predetermined mission is described below, with reference to [Fig.4].

[0068] The method comprises a preliminary step a) of obtaining at least one mission parameter and an optimization criterion. The mission parameter preferably corresponds to a nominal mission of the drone.

[0069] The at least one mission parameter comprises, for example, at least one trajectory, i.e. values ​​of position, speed, orientation and angular rotation speed to be respected at different points of the route. The at least one mission parameter may also comprise a plurality of waypoints as defined above.

[0070] The optimization criterion is, for example, the mass of the aircraft 1 and / or the energy expended by the aircraft 1 during the mission. Depending on the case, these two criteria may prove to be more or less relevant, depending on the duration of the mission, the distances traveled, a possible payload transported by the aircraft 1, etc. The optimization criterion chosen is the one that is considered to be the most crucial, and which will be used to guide the choices of components and the adjustment of the control laws.

[0071] The method may also comprise, in addition to obtaining and providing at least one mission parameter, a step a') of obtaining general parameters not specifically linked to the nominal mission, such as for example a minimum flight speed of the aircraft 1, a maximum flight speed, etc. Such parameters may for example correspond to a flight envelope outside the nominal mission. These general parameters describing such a flight envelope outside the nominal mission may for example comprise speed or acceleration limits, such as maximum and minimum speeds, along the three axes of movement, and / or altitude setpoints along the roll and pitch axes, and / or angular speed setpoints, for example maximum and minimum angular speeds of rotation, around the three axes.

[0072] The method then comprises a step b) of selecting a first combination of components for the aircraft 1. This combination of components constitutes a starting point for the optimization method, and can for example be made by an operator based on his specific knowledge.

[0073] The method then comprises a step c) of obtaining characteristics of each component of the aircraft 1. These characteristics include in particular a mass of the component, as well as one or more specific functional characteristics of the component, such as a length of the arms 6 of a structure 3, a pitch angle, a lift coefficient, a drag coefficient and / or a length of the blades of the rotors 7, nominal and maximum torque values ​​provided by the electric motors 5, an inertia of a motor shaft and an electrical resistance of the motors 5, or even a number of cells and / or a storage capacity of the battery 9.

[0074] The characteristics are advantageously obtained by reading the element identification of each component and a query of a database including all the characteristics of each of the available components. This makes it possible to obtain the characteristics of each of the components quickly and reliably, with a reduced risk of error.

[0075] This reading is advantageously carried out entirely automatically, offline, that is to say without communication between the aircraft computer and an external control device. The identification elements are read by one or more automated devices and the characteristics obtained are directly transmitted to the on-board computer for the calculation of the coefficients of the control laws.

[0076] In the context of a method for determining components having optimal characteristics, this step of obtaining the characteristics of the chosen components makes it possible to determine the real characteristics closest to the calculated optimal characteristics, and to use these real characteristics in the calculation of the control law coefficients.

[0077] The method then comprises a step d) of calculating coefficients of control laws of the aircraft for the combination of components considered.

[0078] As previously indicated, the coefficients of the control laws are the gain factors of the proportional, integral and derivative terms in the feedback algorithms of the different regulated quantities, such as position, speed, angular orientation and angular rotational speed.

[0079] These coefficients are calculated iteratively using an optimization algorithm, seeking to minimize a trajectory tracking error function.

[0080] Said trajectory tracking error function is a function representative of the deviation, for each regulated quantity, between the setpoint and the actual value taken by said quantity over the entire mission. For example, the error function is an integral over the entire path of the deviation between the simulated trajectory and the desired trajectory.

[0081] Alternatively, the square of the deviation can be used to increase the weight of large deviations.

[0082] This error function is advantageously calculated by numerical simulation of the behavior of the aircraft 1 with a given set of control law coefficients, which gives the trajectory followed from the recorded setpoint trajectory.

[0083] The optimization problem is subject to constraints depending on the mission parameters, general parameters and component characteristics.

[0084] In particular, the mission parameters may include constraints on the aircraft's speed, altitude and orientation, or attitude.

[0085] For example, the torque limit values ​​of the motors limit the accessible speed and acceleration ranges.

[0086] External constraints, linked to robustness to external conditions, such as for example taking into account wind or wind gusts, can also be taken into account in the calculation of the coefficients of the control laws.

[0087] Imposing structural constraints on the control laws can make the optimization problem non-convex, and then requires a specific resolution strategy. An example of a particularly effective algorithmic resolution strategy is that based on the generalized gradient method, so-called "bundling" techniques and other non-smooth performance criteria. The details of the operation of non-smooth optimization are developed in the article "Nonsmooth H-infinity sythesis" by P. Apkarian and D. Noll, IEEE 2006.

[0088] The method then comprises a step e) of calculating a value of a cost function for the combination of components considered, in order to determine by an optimization calculation the combination of components minimizing the values ​​of this cost function.

[0089] This cost function is selected from the optimization criterion, i.e. it is for example a function of the total mass of the aircraft or a function of the total energy expended by the propulsion system of the aircraft during the mission and / or a function of a maximum power consumed by the propulsion system of the aircraft during the mission.

[0090] For example, the cost function can be expressed as a linear combination of the total mass of the aircraft with a first coefficient ab, the total energy expended by the propulsion system with a second coefficient a2 and the maximum power consumed with a third coefficient a3.

[0091] The value taken by the cost function is calculated from the characteristics of the components, each mission parameter and the coefficients of the control laws previously obtained.

[0092] The method then comprises a step f), during which a validation criterion is applied to the value of the cost function to determine whether the current combination of components is satisfactory. The validation criterion is for example a threshold value of the cost function, i.e. a maximum value for the cost function for the combination to be satisfactory.

[0093] Alternatively, the validation criterion implements a more complex function of the value of the cost function and / or at least one derivative of the cost function. For example, the validation criterion involves a local minimum of the cost function and a maximum threshold value of the value of the cost function to have a satisfactory combination.

[0094] If the combination is considered satisfactory, i.e. if the validation criterion is met, step f) is stopped and the current combination of components is selected for the aircraft.

[0095] Alternatively, if the selection criterion is not met, step f) then comprises a selection of a new combination of components by modifying at least one of the components of the aircraft 1, in order to reduce the values ​​of the cost function. The alternative component(s) are selected from a database grouping the different components available for the aircraft 1.

[0096] The preceding steps c), d), e) of obtaining the characteristics, calculating the coefficients of the control laws and calculating the values ​​of the cost function are iterated according to an optimization algorithm aiming to minimize the values ​​of the cost function as a function of the sizes of the motors 5, the radii of the rotors 7, the fictitious collective pitch of the rotor 7 and the length of the arms 6.

[0097] The optimization is accompanied by constraints linked to the characteristics of the components.

[0098] For example, a maximum engine torque value Cmotmax may be imposed as an upper limit of the engine torque Cmot for each motor 5.

[0099] Also, a maximum temperature value of the engines n,.,x can be imposed as an upper limit of the temperature ❖mot reached by each engine 5 of the aircraft during the mission.

[0100] Furthermore, a maximum value of rotation speed of the blades corotmax may be imposed as an upper limit for the rotation speed corot of the blades of each rotor 7. This maximum value is for example calculated from the maximum permissible speed for the ends of the blades, depending on their length and their resistance.

[0101] Other constraints may be related to the rotors 7 and the support arms 6, as well as to the compatibility of the components with each other. For example, the diameters of the rotors and the lengths of the arms are linked by a constraint which requires that the distance between the centers of two rotors be greater than the sum of the lengths of the blades, to avoid overlapping.

[0102] Another constraint related to the arms 6 and the rotors 7 may be that the bending resonance frequency of the arm 6 must preferably be higher than the cut-off frequency of the associated motor, and in particular 10 times higher. This makes it possible to prevent the actuation of the motor from causing resonance in the arm and damaging it.

[0103] Other constraints may be linked to the electrical power supplied by the battery 9.

[0104] For example, a constraint may be that the maximum power Pbatmax that can be supplied by the battery must be greater at all times than the sum of the instantaneous powers P mot(t) consumed by the engines and by the other systems of the aircraft.

[0105] Another constraint linked to the battery may be that the total energy Ebat storable in the battery 9 must be greater than the sum of the energies Emot consumed by each engines 5 and by the other systems of aircraft 1 over the entire duration of the mission.

[0106] Preferably, a safety margin of 20% on the energy stored in the Ebat battery is taken, in order to improve the safety of the aircraft.

[0107] The maximum power Pbat and the total energy Ebat generally vary linearly with the mass of the battery 9.

[0108] This nonlinear optimization problem under constraints can for example be solved by substitution models, which make it possible to reduce the number of evaluations of criteria and constraints. Indeed, substitution models are constructed to replace as faithfully as possible the behavior of the simulation model. This construction is done from the simulation model with a limited number of judiciously selected design points.

[0109] Optimization by minimization of the cost function leads to a step g) of obtaining an optimal combination of components of the aircraft 1 for the mission, as well as a set of coefficients of the control laws for this optimal combination.

[0110] The method then comprises a step h) of installing on the aircraft 1 the components according to the determined optimal combination and of transmitting the coefficients of the control laws to the flight controller of the aircraft 1.

[0111] The method of preparing the aircraft 1 described above can be part of a method of carrying out a mission by an aircraft 1, comprising said preparation of the aircraft 1 and at least one step of automatic flight of the aircraft 1 according to the pre-recorded trajectory and with the coefficients of the determined control laws.

Claims

Claims

1. Method for preparing a remotely piloted multi-rotor aircraft (1) for a mission, the method comprising steps of: a) obtaining at least one mission parameter and an optimization criterion, b) selecting a combination of components (3, 5, 7, 9) for the aircraft (1), c) obtaining characteristics of each component (3, 5, 7, 9), d) calculating coefficients of control laws of the aircraft (1) for the combination of components (3, 5, 6, 7), by minimizing a function representative of a trajectory tracking error, from each mission parameter and the characteristics of the components (3, 5, 7, 9), e) calculating a value of a cost function for the combination, said cost function being selected from the optimization criterion, from characteristics of the components (3, 5, 7, 9), from each mission parameter and the coefficients of the control laws,f) applying a validation criterion to the value of the calculated cost function, and, if the validation criterion is not satisfied, selecting a new combination by modifying at least one of the components (3, 5, 7, 9) so as to minimize the value of the cost function, and iterating steps c) to e), g) obtaining an optimal combination of components of the aircraft (1) for the mission and coefficients of the control laws for the optimal combination, and h) installing on the aircraft (1) the components (3, 5, 7, 9) according to the optimal combination and transmitting the coefficients of the control laws to control electronics (15) of the aircraft (1).,

2. Method according to the preceding claim, in which the method also comprises a step a') of obtaining general parameters relating to the aircraft (1) and independent of the mission, said general parameters being taken into account in the choice of the components (3, 5, 7, 9) to minimize the values ​​of the cost function, said general parameters preferably comprising a nominal off-mission flight envelope.

3. Method according to one of the preceding claims, in which step c) of obtaining characteristics comprises reading an identification element of each component (3, 5, 7, 9) and an interrogation of a database comprising characteristics of components (3, 5, 7, 9).

4. Method according to the preceding claim, in which the identification element is a barcode or, preferably, a QR code.

5. Method according to one of the preceding claims, wherein step c) of obtaining the characteristics comprises obtaining at least one of a mass of the component and one or more specific functional characteristics of the component, such as an arm length (6) of a structure (3), a pitch angle, a lift coefficient, a drag coefficient and / or a rotor blade length (7), nominal and maximum torque values ​​provided by motors (5), an inertia of a motor shaft and an electrical resistance of the motors (5), or a number of cells and / or a storage capacity of a battery (9).

6. Method according to one of the preceding claims, in which the cost function is a function of a total mass of the aircraft (1), and / or a function of a total energy expended by a propulsion system of the aircraft (1) during the mission, and / or a function of a maximum power consumed by the propulsion system of the aircraft (1) during the mission.

7. Method according to claim 6, in which the cost function is expressed as a linear combination of the total mass of the aircraft with a first coefficient (ai), the total energy expended by the propulsion system with a second coefficient (a2) and the maximum power consumed with a third coefficient (a3).

8. Method according to one of the preceding claims, in which step d) of calculating coefficients of the control laws comprises taking into account at least one external constraint linked to external conditions.

9. Method according to one of the preceding claims, in which, during step e), the value of the cost function is calculated from a simulation of a flight of the aircraft (1) with the control law coefficients obtained in step d).

10. Method for carrying out a mission by an aircraft (1), comprising the implementation of a method for preparing the aircraft (1) according to one of the preceding claims and a step of flying the aircraft (1) with the components (3, 5, 7, 9) according to the determined optimal combination and provided with the associated control law coefficients.

Images