Method and control unit for controlling an over-determined system, system and aircraft

Abstract

A method for controlling an overdetermined system having multiple actuators, such as an aircraft (1) having multiple drive units (3), the actuators performing at least one primary task and at least one non-primary task, the method comprising: a) determining pseudo-control commands based on a physical model of the system, the pseudo-control commands specifically representing torques (L, M, N) and total thrust (F) acting on the system; b) based on u p =D u Determine the control matrix D, so that u1 = D ‑1 u p , represents the control instructions for the actuator to perform the main task; c) map the non-main task into the null space N(D) of the main task, so that when u2, represents the control instructions for the actuator to implement the non-main task, Du2 = 0, and d) provide the control instructions from b) and c) to the actuator.

Description

Technical Field

[0001] The present invention relates to a method for controlling an overstability system having multiple actuators, such as an aircraft having multiple drive units.

[0002] The present invention also relates to a control unit for controlling an overstability system having multiple actuators, such as an aircraft having multiple drive units.

[0003] Furthermore, the present invention relates to an overstability system having multiple actuators.

[0004] Finally, the present invention relates to an aircraft having multiple drive units and optionally additional actuators, such as, for example, movable flaps or winches, the drive units and actuators together forming an overstability system. Background Technology

[0005] Over-actuation (i.e., an aircraft having more actuators than are needed to perform the so-called primary task in its control space) aircraft, flying machines, or aircraft offers the possibility of achieving the task with more than one solution. This intuitively suggests that, depending on the aircraft type, actuator type, and the task defined in the control space, there may be additional space for performing one or more other (non-primary) tasks.

[0006] Different tasks with corresponding priorities are combined to ensure that lower-priority tasks do not affect higher-priority tasks, which are considered important. The way and form secondary or tertiary (fourth or generally non-essential) tasks are added to the control and configuration problem (assignment problem) can significantly affect the performance available to solve the primary task and the stability of the solution.

[0007] The applicant has proposed a method for operating an eVTOL aircraft with 18 rotors in European patent application 19,212,935.1, in which the main control allocation yields a solution for providing the required thrust and torque using a so-called "L-2" minimization, while secondary or lower-priority (non-primary) constraints complete the reduction of the maximum LTU (Lift-Trust-Unit) power. This is achieved by changing the amplification parameter of the allocation problem as a function of the deviation of the LTU value (speed-RPM) from the average of the total LTU (so-called "L-inf" minimization). Here, two tasks are performed in one step, making it impossible to assign different priorities or different levels of urgency to different objectives. Furthermore, the characteristic of failure or malfunction in achieving objectives with lower urgency (e.g., secondary objectives) may adversely affect the completion of tasks with higher urgency (e.g., primary objectives). Summary of the Invention

[0008] The object of the present invention is to provide a remedy and to give a method, a control unit, a system and an aircraft of the corresponding type, wherein, in addition to performing at least one primary task, at least one non-primary task can be solved without adversely affecting the primary task or the solution to the primary task.

[0009] According to the present invention, a method for controlling a superstitious system having multiple actuators, such as an aircraft having multiple drive units, wherein the actuators perform at least one primary task and at least one non-primary task, the method comprising:

[0010] a) Determine pseudo-control commands based on the physical model of the system. The pseudo-control commands specifically represent the torque and total thrust acting on the system.

[0011] b) According to u p =Du determines the control matrix D, Therefore, u1 = D -1 u p , This refers to the control commands used by the actuator to perform the main task.

[0012] c) Map the non-primary tasks to the null space N(D) of the primary task, so that when u2, When indicating control commands used by the actuator to perform non-primary tasks, Du2 = 0, and

[0013] d) Provide control commands from b) and c) to the actuator.

[0014] According to the present invention, a control unit for controlling a superstitious system having multiple actuators, such as an aircraft having multiple drive units, wherein the actuators perform at least one primary task and at least one non-primary task, and the control unit is specifically constructed and configured in software technology for:

[0015] a) Determine pseudo-control commands based on the physical model of the system. The pseudo-control commands specifically represent the torque and total thrust acting on the system.

[0016] b) According to u p =Du determines the control matrix D, Therefore, u1 = D -1 u p , This refers to the control commands used by the actuator to perform the main task.

[0017] c) Map the non-primary tasks to the null space N(D) of the primary task, so that when u2, When indicating control commands used by the actuator to perform non-primary tasks, Du2 = 0, and

[0018] d) Provide control commands from b) and c) to the actuator so as to manipulate the actuator accordingly.

[0019] The control unit may be further constructed and configured to perform an improved version of the method according to the invention described below.

[0020] The overstability system according to the invention includes a plurality of actuators that are operatively connected to the control unit according to the invention.

[0021] The aircraft according to the invention includes multiple drive units and optionally additional actuators, such as movable flaps or winches, which together constitute an overstability system according to the invention, particularly an aircraft in the form of a multi-rotor aircraft having multiple preferably electrically driven rotor units.

[0022] In the present context, "actuator" specifically refers to, but is not limited to, drive units such as rotor units or the like. Depending on the specific design of the system (aircraft), there may also be other mechanisms, such as flaps, winches, or, in general, the operating mechanisms required by the system to perform specific physical tasks.

[0023] A method for assigning task priorities to overdetermined systems (e.g., aircraft) is proposed, in which low-priority tasks (non-primary tasks) are mapped onto the null space of higher-priority tasks (primary tasks). This ensures that low-priority tasks do not adversely affect high-priority tasks or their solutions.

[0024] In the current context, "null space" should refer to the kernel of a (linear) mathematical mapping. The kernel of a linear mapping f: V → W between vector spaces V and W consists of vectors that map from the zero vector in V to the zero vector in W; that is, the kernel is the set of solutions to the homogeneous linear equation f(x) = 0 and is therefore also called the null space.

[0025] If we consider the example in EP 19 212 935.1, the primary task is to obtain a solution for the control allocation used to generate the necessary (total) thrust and necessary torque using "L-2 minimization". Correspondingly, a secondary task, i.e., a non-primary task, could be to reduce the maximum LTU power using "L-inf minimization". If the secondary task is mapped into the null space of the primary task according to the invention, it is ensured that the secondary task does not affect the solution of the primary task. This is desirable for safety reasons, as the primary task is mandatory, while the secondary task is only preferably to be completed. Furthermore, additional subordinate tasks (e.g., a third task) can be specified, for example, by reducing the power of determined (or multiple) LTUs to, for example, 75% because the relevant drive unit overheats or is classified as faulty. According to a corresponding improvement of the invention, the third task is mapped into the null space of the primary and secondary tasks, so that the third task does not adversely affect the higher-priority task.

[0026] In the example above, the order of secondary and tertiary tasks can be changed depending on specific requirements and priorities. In this application, the required thrust and torque are considered primary tasks, and all others (secondary, tertiary, and other tasks) are considered non-primary.

[0027] If a corresponding location still exists in the null space, the proposed method can also be used for the fourth, fifth, or other tasks. If no location exists in the null space, these tasks are simply ignored based on the definition of the null space.

[0028] It is important to emphasize that the proposed concept offers a highly advantageous differentiation based on the urgency of the mission, which can be used to allocate different Design Assurance Levels (DALs) based on varying levels of urgency and objectives. This has a significant impact on reducing the development costs of safety-critical equipment, such as aircraft.

[0029] Although aircraft are repeatedly referenced here and below, the method described is also fully applicable in general to controlling any type of actuator system.

[0030] For ease of understanding, the invention, particularly the mathematical and physical background of the aircraft, will be briefly and thoroughly explained here.

[0031] The system's equations of motion, which can be derived using Newton's and Euler's theorems or the Lagrange method, can be expressed as follows:

[0032]

[0033] in, It is the system's configuration vector, such as position and rotation in 3D. It is the state-dependent generalized moment of inertia. The Coriolis force represents the state-related force. Represents gravity. Rather, it is caused by external forces and torques, such as those resulting from aerodynamics and contact. The physical control commands (or pseudo-control commands) required for this system are called... The control commands are calculated using, for example, feedback control rules and are used to control the system. The pseudo-control commands are forces and torques fixed relative to the main body, which act on the system through different actuators, and these forces and torques are respectively controlled by a control input matrix. Substitute these into the system dynamics given in Equation 1. These matrices specifically contain over / under-actuation.

[0034] To calculate u p Using control methods (or rules) (e.g., direct correlation or feedback control rules, etc.). These calculated control commands and actual actuator control commands. The correlation between them is accomplished through allocation or configuration issues, specifically concerning the geometry of actuator positioning within the system and other actuator-related configurations and characteristics. This applies here as follows:

[0035] u p =Du, Equation 2

[0036] in Define the so-called control effectiveness matrix (or simply control matrix). As mentioned above, after using the control method, the pseudo-control command u is first calculated. p However, the pseudo-control command must be assigned to the physical actuator in the form of an actual control command u, which is generally known as a control assignment or configuration problem. Therefore, an inverse matrix calculation is required so that u p Calculate u. This can be expressed as:

[0037] u=D -1 (W,u min ,u max )u p Equation 3

[0038] This inverse is typically performed when considering weights or weighted matrices. And under the physical limits of each actuator, for example and in,

[0039] Will It is called the "control allocation matrix" (or configuration matrix).

[0040] For aircraft, such as the 18-rotor aircraft produced by the applicant. This represents a vector containing the required (control) thrust and three-dimensional (operating) torque, which act on the aircraft about its center of gravity. Furthermore, It is a vector containing eighteen (18) desired rotor control commands.

[0041] The previous step used the allocation matrix The distributed actuator control command u is calculated, and the allocation matrix is ​​the result of a matrix inversion problem. For a system with a redundant number of actuators (i.e., k>p'), if the control matrix has a sufficient number of linearly dependent columns (as is known to those skilled in linear algebra), there may be more than one solution to the inversion problem. In this application, such a system is referred to as an "overdetermined system".

[0042] The existence of more than one solution to the inversion problem provides the possibility of a "null space" other than zero, which can be used for other (minor, tertiary, or generally lower priority) tasks without threatening the primary objective (the primary task or its solution). In the current case, the primary task lies in the required physical control command u. p Calculate the actuator control command u.

[0043] A matrix The null space, or kernel, of matrix D is the set of vectors v that make Dv = 0. Therefore, the null space of matrix D is defined as: N(D) = {v|Dv = 0}.

[0044] One possible representation of the null space of a matrix can be calculated using the following formula:

[0045]

[0046] in, It is a k×k identity matrix.

[0047] If we assume that the main objective is achieved by using an allocation matrix, that is to say,

[0048] u1 = D -1 u p ,

[0049] but It is a vector containing actuator instructions that generate the necessary physical control instructions u. p The primary goal is to solve the main task. This allocation setup produces executable instructions, that is:

[0050]

[0051] According to a corresponding design scheme of the method of the present invention, based on Calculate the null space N(D), u1=D -1 u p ,in It includes solving the main task and generating the physical control commands u required for the actuator. p Vectors or control commands, here

[0052] If we now examine the calculations made to achieve a secondary (non-primary) objective... Then it can be calculated according to the following formula Mapped onto the null space of the control matrix used for the primary objective:

[0053]

[0054] According to a corresponding design scheme of the method of the present invention, the calculation is first performed. In order to resolve non-primary tasks, and then according to Will It is projected onto the null space N(D) of the control matrix D.

[0055] At this point, the following overall control command is specifically calculated:

[0056] u = u1 + u2.

[0057] In this way, both u and u1 generate the same physical control command u. p ,Right now:

[0058] u p =Du=Du1,

[0059] because

[0060] This ensures that the primary objective is achieved under all circumstances, and that the secondary objective is achieved only when there is sufficient space in the null space.

[0061] According to a corresponding design scheme of the method of the present invention, the total control command is calculated as follows: u = u1 + u2, wherein u and u1 generate the same control command u. p This is because, due to u p =Du=Du1.

[0062] To advantageously ensure the executability of the general instruction, i.e., u∈U, the following limits are preferably observed in the secondary allocation steps:

[0063]

[0064] That is to say, according to a corresponding design scheme of the method according to the present invention,

[0065] The same principle can be applied to the third task (and correspondingly to other subordinate tasks), assuming, for example, that... Then, the following conclusions are drawn:

[0066]

[0067] At this time, the master control command is:

[0068] u = u1 + u2 + u3,

[0069] This also ensures that the primary objective is always achieved, that is to say:

[0070] u p =Du=Du1.

[0071] This process continues recursively, and for the i-th task, it proceeds as follows:

[0072]

[0073] and

[0074]

[0075] Similarly, the generalized extrema can be written as:

[0076]

[0077] in

[0078] This control allocation based on task priority in the use of null space mapping is particularly advantageous for aircraft with distributed actuators.

[0079] As an example of a superdetermined system, consider a multi-motor aircraft with k=18 actuators, which are fixed in a common main reference system, corresponding to the one sold by the applicant. and The former is used as an air taxi to transport passengers, while the latter is used to transport payloads.

[0080] When the system overtimes, the proposed method can also be applied to any other system with a large number of actuators. If these actuators can move about a common host reference system (e.g., pitch actuators or flaps / wings), the control matrix can differentiate between tasks with different priorities. The proposed method can also be used in this configuration.

[0081] The primary task is to provide physical control forces (such as torque and thrust), which generate (flight) control rules, that is:

[0082] u1 = D -1 u p ,

[0083] The solution to the inverse problem (assignment matrix) is preferably calculated such that the solution minimizes the energy consumption caused by u1. That is, A = D. -1 (W,u min ,u max Applying the L-2 minimization objective to u1 T Wu1 (for example, in the form of the known Moore-Penrose pseudoinverse).

[0084] According to a corresponding design scheme of the present invention, for the main task, based on u1 = D -1 u p This is how the inverse of the control matrix D is calculated, i.e., making the expression u1 T Wu1 is minimized, for example, using the Moore-Penrose generalized inverse method, where for each actuator, u = D. -1 (W,u min ,u max )u p Weight matrix And physical load limit as well as in,

[0085] Secondary (non-primary) tasks can be designed to reduce the maximum peak value in actuator instructions, but are certainly not limited to this. However, unlike the case in 19 212 935.1, the current approach ensures that the secondary task, or its solution, does not adversely affect the primary task. That is, the secondary task is implemented by computing an actuator instruction u2 that attempts to bring the components of the primary actuator instruction u1 that are above the average value (or other benchmark value) closer to that average value. This achieves an equivalent distribution of actuator instructions (for completing the primary task) that implicitly reduces the maximum peak value. In this sense, the described secondary task has an L-inf minimization objective.

[0086] Therefore, it is preferable to first calculate the average value of the actuator commands required to achieve the main objective u1:

[0087]

[0088] The difference between the actuator command and the average value used to achieve the main objective is given by the following formula:

[0089] ud =u1-u 1,m .

[0090] At this point, the goal of the secondary computation task is to reduce the amount of data used for the final actuator instruction u. d Actuator commands:

[0091]

[0092] According to a corresponding design scheme of the present invention, the control instructions for solving non-primary tasks are calculated in this way. The component u1 of the control command used to solve the main task approaches its average value through the following steps:

[0093] i) Calculate the average value of the control command u1 used to complete the main task.

[0094]

[0095] ii) Determine the difference between the control command u1 used to complete the main task and the average value:

[0096] u d =u1-u 1,m

[0097] iii) Calculate control instructions for performing non-primary tasks so that u d Decrease:

[0098]

[0099] The null space is calculated as follows:

[0100]

[0101] And the secondary tasks are mapped onto this null space as follows:

[0102]

[0103] A preferred embodiment of the present invention specifies that control commands are also used for the primary task and the non-primary task. Define at least one subordinate task, which is executed by the actuator, for example, for at least one actuator j,μ j Load reduction μ performed in ∈[0,1] j ,thereby

[0104]

[0105] The overall control command is then calculated as follows:

[0106] u = u1 + u2 + u3.

[0107] Alternative locations can be set and attached to both primary and non-primary tasks using control commands. Define at least one (additional) subordinate task, which can be performed by an actuator, for example, for at least one actuator j, μ j ∈[0,1] reduce load μ j ,thereby

[0108]

[0109] The total control command is calculated as follows: u = u1 + u2 + u3.

[0110] This third task is preferably defined so that the instructions for the individual actuators can be adapted when, for example, one or more actuators must perform a task under reduced load. This is the case when one or more actuators, despite being considered, experience overload or overheating. There may also be other reasons for adapting the instructions for the individual actuators.

[0111] As an example, it can be assumed that actuator j should have an efficiency μ j ,μ j It operates within [0,1] (e.g., 0.75, or 75%). If desired, instructions for more than one actuator can also be adapted. The decision of which actuators to operate with what efficiency can preferably be derived through an actuator observation function (implemented in the form of an actuator observation device), the details of which are not discussed in detail in this application.

[0112] At this time, as mentioned above, according to... Figure 1A and Figure 1B The example illustrates the logic used to calculate the actuator instructions that implement the third task.

[0113] The control command D already exists for a system where the actuator is fixed relative to the aircraft. Therefore, for this task, the null space is also N, as given above.

[0114] At this point, the third task is mapped onto the null space so that it does not affect the secondary and primary tasks:

[0115]

[0116] Finally, the overall actuator command (or control command) is expressed as follows:

[0117] u = u1 + u2 + u3,

[0118] This also ensures that the primary task is not affected by secondary and tertiary tasks (non-primary tasks).

[0119] Within the scope of this invention, the order of these tasks can be changed according to priority.

[0120] According to a specific improvement of the control unit of the present invention, the control unit is operatively connected to means for measuring and / or determining the parameters and states of the system and / or actuators, which are required to determine the pseudo-control commands according to step a), particularly the main flight control unit.

[0121] According to another improved embodiment of the control unit of the present invention, the control unit is operatively connected to an actuator observation unit, which is configured and set to determine, based on parameters and states measured or determined by the system and / or the actuators, which actuators j perform the subordinate tasks as described above and, in particular, provide matching values ​​μ when necessary. j ∈[0,1].

[0122] The following description refers specifically to the application scenario of "aircraft". Accordingly, according to an improved embodiment of the method of the present invention, the method is applied to an aircraft having a plurality of control units, preferably 18 control units, i.e., k=18, wherein the control units form at least some actuators of the system.

[0123] The corresponding description introduces a method for controlling a superdeterministic aircraft using mission / objective prioritization, wherein different numbers of tasks of the aircraft's actuators can be managed such that the primary task is not adversely affected by non-primary tasks, and all tasks are accomplished as well as possible in relation to the aircraft design. This is ensured by allocating non-primary tasks to the null space of the primary task, as has been explained several times. The aircraft's structure (e.g., actuator positioning) plays a crucial role in the configuration of the null space; this null space defines the extent to which the primary task can be accomplished when it is not threatened.

[0124] The mathematical principles described herein are general. The algorithmic implementation of these mathematical principles generally does not require external measurement or sensor input. The control matrix D and the primary and secondary control tasks / objectives must be known or predefined in their priority order. This application focuses, for example, on aircraft-type applications. or The implementation of a method for a superdetermined system of this type, wherein the aircraft has a 4-dimensional control space but (non-limitingly) 18 actuators. The following describes how the method or corresponding algorithm can be implemented for such an aircraft.

[0125] Flight control rules are used to calculate pseudo-control commands for achieving stable and controlled flight tracking (e.g., attitude, altitude, and, if possible, position and velocity). The pseudo-control commands represent three torques and one thrust, which should physically act on the main body of the aircraft. This is in Figure 2 The figures are schematically illustrated. Here, L, M, and N represent the torques about the axes x, y, and z (roll, pitch, and yaw axes) of aircraft 1, while F represents the total thrust. Reference numeral 2 indicates the (main) aircraft control unit of aircraft 1, which implements or executes the flight control rules, and reference numeral 3 indicates 18 identical control units (actuators), each comprising a (electric) motor 3a and a rotor 3b. Reference numeral 4 exemplarily indicates one of a plurality of sensor units operatively connected to the flight control unit 2.

[0126] The primary objective (major task) of the flight control unit is to realize the four-dimensional thrust and torque vectors acting on the aircraft (generally acting on or around the aircraft's center of gravity). For this reason, the first step is to calculate u1 = D. -1 u p (in It is a vector containing actuator commands, from which the control commands are calculated to achieve the main objective. The solution to the inverse problem (assignment or configuration matrix A) is preferably calculated such that the solution minimizes energy consumption. This means: A = D -1 (W,u min ,u max ), can utilize u1 T The L-2 minimization objective of Wu1 (e.g., the Moore-Penrose pseudo-inverse method) is performed considering the physical actuator load limits. u should be performed beforehand. p The calculations are preferably performed via master (flight) control rules implemented in the (master) flight control unit, which require or perform predictions of the aircraft's attitude state (3D rotations or rotation matrices in angular or quaternion form and their derivatives, i.e., rotational velocity or rotational acceleration), altitude predictions (vertical altitude, velocity, and acceleration), and, where applicable, horizontal position / velocity and acceleration predictions. To calculate these predictions, conventional methods require inertial instruments, barometers, GNSS (Global Navigation Satellite System), and, if present, cameras / ultrasonic / LiDAR / radar as sensors to improve the quality of the predictions (see...). Figure 2(See attached figure 4). The flight control rules used are therefore typically state-feedback closed-loop control rules, which are designed to stabilize the aircraft and provide sufficient controllability and operability, while being robust to unknown conditions both inside and outside the system.

[0127] As mentioned above, the primary objective is the pseudo-control instruction u. p The pseudo-control command is then sent as control command u1 to each actuator. However, a secondary objective is pre-calculated, and its share is added to the actuator commands without affecting the primary objective (because the null space of the primary task is used for this). The secondary task is performed only during periods when it does not involve the primary task. The secondary task might involve reducing the maximum peak value in the actuator commands, as described in other parts of application EP 19 212935.1, but in this case, it is ensured that the secondary task does not affect the primary task. That is, the secondary task is specifically addressed by calculating a control command u2 that brings the components of the primary actuator command u1 that are above the average value (or other benchmark value) closer to the average value. This technique makes the actuator commands (in the sense of the primary task) equally distributed while implicitly reducing the maximum peak value. In this case, the secondary task is specifically an L-inf-minimization objective. This step does not require sensor data input.

[0128] It should be noted that the main tasks are designed in four-dimensional space, in which the physical torques and thrusts acting on the spacecraft are defined (see...). Figure 2 On the other hand, the secondary task is first designed in the current 18-dimensional space, in which actuator instructions are defined.

[0129] To achieve this, we first calculate the average value of the actuator commands that achieve the main objective u1:

[0130]

[0131] Then calculate the difference between the actuator command that achieves the main objective and the average value:

[0132] u d =u1-u 1,m .

[0133] Now, actuator instructions for the secondary task are calculated, the objective of which is to reduce u. d This allows us to obtain the final actuator instructions for the secondary task:

[0134]

[0135] This represents the original actuator instructions used only to achieve secondary objectives. To ensure that secondary tasks do not affect primary tasks, the null space of the control matrix is ​​first calculated, and then... This is mapped onto the null space. The null space is calculated as follows:

[0136]

[0137] Then, the secondary tasks are mapped onto the null space as follows:

[0138]

[0139] Therefore, it is ensured that the relevant control instructions achieve the main objective (i.e., obtain u) when u = u1 + u2 and when u = u1. p ), and additionally u = u1 + u2 makes it possible to achieve the secondary objective (provided the null space allows it).

[0140] Using the logic described, the control scope can include other tasks / objectives with lower priority. For example, when one or more actuators must bear a smaller load for a defined reason, the instructions for the individual actuators can be adapted as an additional (third) task selection. This may occur when, for example, an overload, overheating, or other malfunction of one or more actuators should be considered. However, there may also be other reasons for adapting the instructions for the individual actuators. The logic capable of identifying such a state is not part of this application: but the corresponding method can employ available aircraft conditions (in the case of using an inertial measurement unit, GNSS, barometer, etc.), vibration sensors on the actuators, temperature sensors on the actuators, and actuator and aircraft models (see [link to application]). Figure 1A , Figure 1B and 2 (Ref. 4 in the attached figure). In this application, the relevant logic is referred to as an "actuator observation device," and the logic / algorithm preferably provides at least two important values: which actuators (possibly multiple actuators) should be adapted and adjusted (u j And to what extent (μ) should they be respectively j Adapt and adjust the values ​​∈[0,1]). Then, optimize the implementation or execution. Figure 1A and Figure 1B One of the logics shown is used to calculate

[0141] The third task, or its solution, thus determined, is mapped onto the null space as follows:

[0142]

[0143] In this case, the final master control command sent to each actuator is:

[0144] u = u1 + u2 + u3.

[0145] It should also be noted that both u = u1 + u2 + u3 and u = u1 can ensure that the main objective is achieved (providing u) p The first control instruction mentioned, u = u1 + u2 + u3, is additionally used to achieve secondary and tertiary objectives (if the null space allows it). Attached Figure Description

[0146] Other features and advantages of the present invention will become apparent from the following description of the accompanying drawings.

[0147] Figure 1A This illustrates a logic for an actuator observation function or device used in the method provided by the present invention;

[0148] Figure 1B This illustrates another logic for the actuator observation function or device used in the method provided by the present invention;

[0149] Figure 2 An aircraft according to the present invention is shown;

[0150] Figure 3 One possible process of the method according to the present invention is shown. Detailed Implementation

[0151] exist Figure 1A and Figure 1B In the accompanying drawing, reference numeral 2a shows a specific improvement of the control unit for a superstitious system according to the invention, such as an aircraft (see also...). Figure 2 (and related explanations). This improved scheme sets the control unit 2a in operative connection with a device for measuring and / or determining the parameters and states of the system and / or actuators, which are required to determine pseudo-control commands, particularly the main flight control unit, according to step a). The device (sensor) is symbolically shown at reference numeral 4 in the attached drawing (see also...). Figure 2 (and related explanations).

[0152] The control unit 2a is operatively connected to the actuator observation device 2a, which is configured and set to determine, as described above, which actuators j should perform the next-level (third) task, and especially to provide the corresponding value μ when necessary, based on the system and / or actuator parameters and states measured or determined at reference numeral 4. j ∈[0,1], and then through the actual control algorithm (see also) Figure 2 (and related explanations) Use the values ​​stated therein.

[0153] Figure 1A This illustrates a situation where control instructions are also used for both primary and non-primary tasks. At least one (additional) subordinate task is defined, which is performed by an actuator, for example, for at least one actuator j,μ. j The load decreases by μ in [0,1]. j ,thereby

[0154]

[0155] The overall control command is then calculated as follows: u = u1 + u2 + u3. For actuator j, the load reduction factor μ is... j Apply to the (primary and secondary) control instructions u1(j)+u2(j) to obtain a new (third) control instruction based on u2.

[0156] Alternative locations can be set, such as in Figure 1B As shown, control commands are also used for both primary and non-primary tasks. At least one (additional) subordinate task is defined, which is performed by an actuator, for example, for at least one actuator j,μ. j The load decreases by μ in [0,1]. j ,thereby

[0157]

[0158] The overall control command is then calculated as follows: u = u1 + u2 + u3. The load reduction factor μ is then applied. j Applied to the maximum (physical) control value u of actuator j max So that based on u max Obtain new (third) control commands

[0159] Preferably, this third task can be defined so that, for example, instructions for adjusting the individual actuators can be adapted when one or more actuators must perform a task under reduced load. This is the case when one or more actuators are overloaded or overheated, although the one or more actuators are taken into consideration.

[0160] As an example, it is assumed that actuator j should have an efficiency μ j ,μ j The actuators operate within the range of [0,1] (e.g., 0.75, or 75%). The decision of which actuators to operate with what efficiency can be preferably made through the actuator observation function (implemented in the form of actuator observation device 2a").

[0161] As already mentioned, Figure 2The system according to the invention, or an aircraft 1 having 18 drive units (actuators), is shown. Figure 2 In the figure, L, M, and N represent the torques about the axes x, y, and z (roll axis, pitch axis, and yaw axis) of the aircraft 1, while F represents the total thrust. Reference numeral 2 symbolically represents the main flight control unit of the aircraft 1, which preferably includes control algorithm 2a' and... Figure 1A and Figure 1B The actuator observation device 2a” is generally configured to perform the method and its improvements according to the invention, particularly in terms of software technology. A human pilot is additionally shown at reference numeral 2b, but this is not of further concern to the human pilot in the present case. Reference numeral 3 indicates 18 (non-limitingly identical) control units (actuators), each comprising a (electric) motor 3a and a rotor 3b. Reference numeral 4 exemplarily indicates one of the sensor units operatively connected to the main flight control unit 2, as described above. Multiple such sensor units 4 may be provided to accommodate the available aircraft states in improvements to the method according to the invention, particularly inertial instruments, GNSS, barometers, vibration sensors on actuators, temperature sensors on actuators, and the like.

[0162] The invention is not limited to use as a superstitious system for aircraft 1.

[0163] Figure 3 The following is a possible flow of the method according to the invention. The method begins in step S1. In step S2, as detailed above, a pseudo-control command is calculated. For this purpose, the (sensor) measurement value from step S2' can be used, as also explained. In step S3, u1 = D is calculated using the control matrix D. -1 u p This is to address the primary task, as already explained. In step S4, secondary tasks are defined and the relevant control instructions are calculated. For this purpose, measurements from the sensor, such as temperature measurements, can be used, as previously explained. Then, in step S5, the non-primary (secondary) task or corresponding control command is transferred. Mapped into the null space N(D) of the main task, so that when u2, When indicating control commands for the actuator used to perform non-primary tasks, Du2 = 0. This has also been explained in detail above.

[0164] Step S6 includes a query to determine whether additional subordinate (non-primary, e.g., third) tasks should be addressed. If the answer is "Yes(j)", then steps S4 (and S4' if necessary) and S5 are adapted (e.g., using...). The steps are repeated for the other tasks. Once the response to the query in step S6 is "No (n)," the method jumps to step S7, where a general control command is determined: = u1 + u2 + ..., depending on the specific number of tasks to be solved (see above). In step S8, this general control command is used to control the actuator, and the method ends in step S9.

Claims

1. A method for controlling a superstitious system having multiple actuators, said system being an aircraft (1) having multiple drive units (3), said actuators performing at least one primary task and at least one non-primary task, said method comprising: a) Determine pseudo-control commands based on the physical model of the system. The pseudo-control commands represent the torques L, M, N, and total thrust F acting on the system, wherein the torques L, M, and N represent the torques about the x-axis, y-axis, and z-axis of the aircraft (1), respectively. b) According to Determine the control matrix , ,thereby , This refers to the control commands used by the actuator to perform the main task. c) Map non-primary tasks to the null space of the primary task. In the middle, thus when , When indicating control commands used by the actuator to perform non-primary tasks, ,as well as d) Provide control commands from b) and c) to the actuator.

2. The method according to claim 1, wherein, according to Calculate the null space ,here ,in This includes solving the main tasks and generating the physical control commands needed for the actuators. Vectors or control commands, It is a k×k identity matrix, where: 。 3. The method according to claim 1, wherein, First calculate In order to resolve non-primary tasks, and then according to Will Projected onto the control matrix zero space Above, among which, This represents the original actuator instruction used only to achieve secondary objectives.

4. The method according to claim 3, wherein, The following steps calculate the overall control command. : ， in, and Generate the same control commands .

5. The method according to claim 3, wherein, for Regulation, 。 6. The method according to claim 1, wherein the method is applied to an aircraft (1) having a plurality of drive units (3) constituting at least some actuators of the system.

7. The method according to any one of claims 1 to 6, wherein, For the main tasks This is how the inverse of the control matrix D is calculated, i.e., using the Moore-Penrose generalized inverse method, so that the expression... Minimize, where for each actuator there is Where W is the weight matrix, and in the case of multiple weight matrices, the i-th weight matrix is ​​defined as... and physical load limit as well as ,in, .

8. The method according to claim 7 when referring to claim 3, wherein, This is how control instructions used to handle non-critical tasks are calculated. This makes the components of the control commands used to solve the main task... Approximate the average value by following these steps: i) Calculate the control instructions used to accomplish the main tasks. average : ， ii) Determine the control instructions used to solve the main task. Difference from the average : ， iii) Calculate control instructions for resolving non-primary tasks, so that... Decrease: , 。 9. The method according to claim 1, wherein, In addition to the primary task and the non-primary task, control commands are also used. Define at least one subordinate task, which is executed by the actuator, wherein the subordinate task is for at least one actuator. j Reduce load , Therefore: , The general control command is then calculated as follows: 。 10. The method according to claim 1, wherein, In addition to the primary task and the non-primary task, control commands are also used. Define at least one subordinate task, which is executed by the actuator, wherein the subordinate task is for at least one actuator. j Reduce load , Therefore, , The overall control command is then calculated as follows: 。 11. A control unit (2a) for controlling an overstability system having multiple actuators, the system being, for example, an aircraft (1) having multiple drive units (3), the actuators performing at least one primary task and at least one non-primary task, the control unit being constructed and configured, particularly in software technology, for: a) Determine pseudo-control commands based on the physical model of the system. The pseudo-control commands specifically represent the torques L, M, N, and total thrust F acting on the system, wherein, The torques L, M, and N represent the torques about the x, y, and z axes of the aircraft (1), respectively. b) According to Determine the control matrix , ,thereby , This refers to the control commands used by the actuator to perform the main task. c) Map non-primary tasks to the null space of the primary task. In the middle, thus when , When indicating control commands used by the actuator to perform non-primary tasks, ,as well as d) Provide control commands from b) and c) to the actuator so as to manipulate the actuator accordingly.

12. The control unit (2a) according to claim 11, wherein the control unit is configured and set to perform the method according to claim 2.

13. The control unit (2a) according to claim 11, wherein the control unit is operatively connected to the device (4), the device (4) being used to perform: measuring the parameters and state of the system, measuring the parameters and state of the actuator, determining the parameters and state of the system, and determining the parameters and state of the actuator, wherein the parameters and state are required to determine the pseudo-control command according to step a).

14. The control unit (2a) according to claim 11, wherein the control unit is operatively connected to an actuator observation unit, the actuator observation unit being configured and set to: determine, based on parameters and states measured or determined by the system and / or the actuator, to perform the subordinate task according to claim 9 or 10 and to provide matching values ​​when necessary. actuator j .

15. An overstability system having multiple actuators, said actuators being operatively connected to a control unit (2a) according to claim 11.

16. An aircraft (1) having multiple drive units (3) and other actuators, the drive units and actuators together forming a hyperstability system according to claim 15, the aircraft being a multi-rotor aircraft having multiple electrically driven rotor units.

Images