A method and system for controlling the flight attitude of a flight simulator

Abstract

The application discloses a flight attitude control method and system of a flight simulator, and belongs to the technical field of fault prediction and health management. Motion state signals of a simulated airplane, a motion platform pose feedback signal and each actuator stroke feedback signal are acquired, and an attitude angle limit allowance and an actuator remaining stroke are calculated. An actuator saturation prediction point is predicted within a preset prediction time domain, and the motion state signals are filtered according to the actuator saturation prediction point, so that a high-frequency motion clue component and a filtered return component are obtained. After phase lead compensation is performed on the filtered return component, the filtered return component is combined with the high-frequency motion clue component, so that a motion platform expected pose instruction is generated, and then each actuator target stroke is solved and controlled. The application can identify an actuator saturation risk in advance, release a platform motion space, reduce attitude top cutting and reverse impact, and improve the authenticity and continuity of flight attitude simulation.

Description

Technical Field

[0001] This invention relates to the field of fault prediction and health management technology, specifically to a method and system for controlling the flight attitude of a flight simulator. Background Technology

[0002] Flight simulators typically generate motion state signals such as pitch, roll, and yaw of the aircraft during training using flight dynamics models. These motion state signals are then converted into attitude commands that a six-degree-of-freedom motion platform can execute, allowing the pilot to obtain attitude perception close to that of real flight. Since the attitude angles, actuator strokes, and speeds of the motion platform are physically limited, existing methods generally employ fixed-parameter filtering to ensure the platform gradually returns to its intermediate position after outputting transient attitudes, thus preventing the platform from continuously approaching its limits.

[0003] However, when performing training exercises such as low-altitude go-around, crosswind landing recovery, or single-engine failure correction in flight simulators, the attitude changes output by the flight dynamics model often exhibit characteristics of continuous reversal, short-duration large amplitude, and multi-axis coupling. For example, during a crosswind landing recovery, the pilot needs to first perform a large roll correction, then quickly reverse the bank angle and coordinate with a pitch pull-up. At this time, the motion platform needs to generate attitude cues according to the target motion command, and at the same time, it needs to release the occupied actuator travel according to the washout and return-to-center command. When the target motion command and the washout and return-to-center command are in opposite directions, and some actuators are close to the lower limit of their remaining travel, fixed washout parameters will cause the platform's return-to-center action and the new attitude cues to cancel each other out or overlap, resulting in short-term hysteresis, reverse impact, or attitude shearing on the platform.

[0004] The aforementioned phenomena are not obvious during routine, stable flight training. However, in low-altitude complex weather or abnormal situation handling training, pilots may perceive pitch or roll changes inconsistent with visual and instrument readings. In severe cases, this can lead to pilots misjudging the aircraft's recovery trend, reducing the realism and safety of the training. Therefore, how to identify actuator saturation risks in advance and compensate for the recovery component when the workspace of the motion platform is limited and the recovery maneuver conflicts with the target attitude maneuver has become an urgent technical problem to be solved. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for controlling the flight attitude of a flight simulator, so as to overcome the shortcomings of the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for controlling the flight attitude of a flight simulator, comprising:

[0007] The simulated aircraft motion state signal output by the flight dynamics model is acquired, as well as the attitude feedback signal of the motion platform and the stroke feedback signal of each actuator; wherein, the simulated aircraft motion state signal includes pitch angle, roll angle, yaw angle, angular velocity and angular acceleration;

[0008] Based on the posture feedback signal and the stroke feedback signal of each actuator, the workspace constraint parameters of the motion platform are calculated. The workspace constraint parameters include the posture angle limit margin and the remaining stroke of the actuator.

[0009] Based on the simulated aircraft motion state signal and workspace constraint parameters, rolling prediction is performed within a preset prediction time domain to obtain the actuator saturation prediction point. Based on the actuator saturation prediction point, the simulated aircraft motion state signal is subjected to washout filtering to obtain high-frequency motion clue components and washout return components.

[0010] Phase advance compensation is performed on the wash-out return component, and it is synthesized with the high-frequency motion clue component to obtain the desired pose command of the motion platform. The target stroke of each actuator is solved according to the desired pose command of the motion platform, and the action of each actuator is controlled.

[0011] Preferably, the calculation of the attitude angle limit margin includes:

[0012] The current pose represented by the pose feedback signal is determined as the reference pose. Preset virtual angle increments are successively superimposed on the reference pose along the positive and negative directions of pitch, roll, and yaw to obtain the corresponding virtual attitude sequence. Inverse kinematics is performed on each virtual attitude in the virtual attitude sequence, and the travel correction amount obtained by subtracting the theoretical travel under the current pose from the travel feedback value is superimposed on the corresponding theoretical virtual travel to obtain the corrected virtual travel. When any corrected virtual travel first reaches the preset travel boundary of the corresponding actuator, the difference between the corresponding virtual attitude angle and the current attitude angle is determined as the attitude angle limit margin in the corresponding direction.

[0013] Preferably, the calculation of the attitude angle limit margin also includes:

[0014] The virtual attitude at which the corrected virtual travel first reaches the preset travel boundary is combined with the previous virtual attitude that did not reach the preset travel boundary to form an angle search interval; inverse kinematics is solved using the midpoint attitude of the angle search interval, and the sub-interval containing the travel boundary is retained based on whether the obtained corrected virtual travel reaches the preset travel boundary; the angle search interval is repeatedly reduced until the interval width is no greater than the preset angle accuracy, and the attitude angle on the side that did not reach the preset travel boundary is determined as the attitude limit angle.

[0015] Preferably, the calculation of the remaining stroke of the actuator includes:

[0016] Based on the pose feedback signal, the motion platform is solved by inverse kinematics to obtain the theoretical stroke of each actuator in the current pose. The corresponding stroke feedback value is then subtracted from the theoretical stroke to obtain the stroke correction amount of each actuator.

[0017] The stroke correction amount is superimposed on the preset minimum stroke boundary and preset maximum stroke boundary of the corresponding actuator to obtain the equivalent minimum stroke boundary and equivalent maximum stroke boundary;

[0018] The extension and retraction direction of each actuator is determined based on the simulated aircraft motion state signal, and the remaining stroke of each actuator is obtained by subtracting the stroke feedback value from the corresponding equivalent stroke boundary according to the extension and retraction direction.

[0019] Preferably, rolling prediction is performed within a preset prediction time domain to obtain the actuator saturation prediction point, including:

[0020] Using the current attitude angle, angular velocity, and angular acceleration as the initial state, the attitude change sequence in the preset prediction time domain is obtained by recursion according to the control cycle, and the attitude change sequence is superimposed on the current pose of the motion platform to obtain the predicted attitude sequence.

[0021] Inverse kinematics is performed on each predicted posture to obtain the predicted stroke of each actuator. The corresponding remaining stroke is selected according to the predicted extension and retraction direction, and the stroke occupancy rate of the predicted stroke change to the remaining stroke is calculated.

[0022] Based on the prediction time from near to far, the prediction time when the stroke occupancy rate of any actuator first reaches the preset saturation value, the actuator number, and the predicted extension / retraction direction are determined as the actuator saturation prediction point.

[0023] Preferably, the washing and filtering process includes:

[0024] Based on the predicted stroke change and remaining stroke corresponding to the actuator saturation prediction point, the predicted remaining stroke is obtained, and the saturation time distance is obtained based on the predicted time of the actuator saturation prediction point. Based on the ratio of the predicted remaining stroke to the remaining stroke and the ratio of the saturation time distance to the preset prediction time domain length, the cutoff adjustment coefficient is determined, and the preset base cutoff frequency is adjusted using the cutoff adjustment coefficient. The equivalent attitude input formed by attitude angle, angular velocity and angular acceleration is high-pass filtered according to the adjusted cutoff frequency to obtain the high-frequency motion clue component. The difference between the equivalent attitude input and the high-frequency motion clue component is reversed to obtain the washout return component.

[0025] Preferably, before performing phase lead compensation on the washed-out return component, the method further includes:

[0026] Using the predicted attitude corresponding to the saturation prediction point of the actuator as a reference, positive and negative angle perturbations are applied along the pitch, roll, and yaw directions, respectively. The inverse kinematics of the perturbed attitude is then solved to obtain the stroke sensitivity of the saturation actuator relative to each attitude direction. The stroke sensitivity of each attitude direction, the corresponding washout return component, and the predicted extension / retraction direction of the saturation actuator are correlated with signs to obtain the boundary approach amount corresponding to each washout return component. The washout return component with a negative boundary approach amount and an absolute value greater than a preset judgment threshold is determined as the target return component that drives the saturation actuator away from the corresponding equivalent stroke boundary.

[0027] Preferably, phase lead compensation includes:

[0028] The saturation urgency coefficient is determined based on the ratio of the saturation time distance to the preset prediction time domain length. The lead time constant and lag time constant of the phase lead element are then determined according to the saturation urgency coefficient, ensuring that the lag time constant is less than the lead time constant. The phase lead element, composed of the lead time constant and lag time constant, is discretized according to the control cycle. The discretized phase lead element is then used to filter the target return center component to obtain the target return center component output in advance. The target return center component output in advance is combined with the original washed-out return center component that was not determined as a target return center component to obtain the compensated washed-out return center component.

[0029] Preferably, obtaining the desired pose command of the motion platform and controlling the actions of each actuator includes:

[0030] The compensated washout return component and the high-frequency motion cue component are superimposed onto the preset mid-position attitude of the motion platform in the pitch, roll and yaw directions, and the superposition result is restricted by the attitude angle limit margin in each direction to obtain the desired pose command of the motion platform. The desired pose command of the motion platform is solved by inverse kinematics, and the obtained theoretical target stroke is restricted between the equivalent minimum stroke boundary and the equivalent maximum stroke boundary of the corresponding actuator to obtain the target stroke of each actuator. The target stroke of each actuator is subtracted from the corresponding stroke feedback value, and the actuator extension and retraction control command is generated according to the obtained stroke tracking error to control each actuator to track the target stroke.

[0031] The present invention also provides a control system for the flight attitude of a flight simulator, comprising: a signal acquisition module, a workspace constraint parameter calculation module, a saturation prediction and washout filtering module, and an actuator control module.

[0032] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0033] This invention uses the real-time pose feedback of the motion platform as a reference, successively constructing virtual attitudes along the positive and negative directions of pitch, roll, and yaw. The corresponding virtual actuator stroke is obtained through inverse kinematics, and the deviation between the theoretical stroke and the actual stroke feedback is used to correct the virtual stroke and the preset stroke boundary. This yields an attitude angle limit margin and remaining actuator stroke that are compatible with the platform's current assembly errors, transmission backlash, and zero-point drift. Based on this, the simulated aircraft's attitude angles, angular velocities, and angular accelerations are recursively extrapolated, and each predicted attitude is mapped to the actuator's predicted stroke. The moment the preset stroke occupancy rate is first reached is determined as the actuator saturation prediction point. This method ensures that platform constraints are no longer characterized by static nominal parameters, but are jointly determined by the current pose, actual feedback, and future motion trends. It can identify saturation risks before the actuator reaches its physical boundary, reducing limit misjudgments caused by model deviations and providing an accurate triggering basis for subsequent early release of the platform's motion space.

[0034] This invention adjusts the washout filtering parameters based on the predicted remaining travel corresponding to the actuator saturation prediction point and the time distance from the current moment. This enables the simulated aircraft motion state signal to form high-frequency motion cue components and washout return-to-center components that match the currently available workspace. Furthermore, using the predicted attitude of the saturated actuator as a reference, the influence direction of each attitude direction on the actuator travel is calculated through attitude disturbances and inverse kinematics. Only the return-to-center components that can move the saturated actuator away from the corresponding equivalent travel boundary are selected, and phase advance compensation is applied to the selected return-to-center components based on the saturation time distance. Therefore, the platform can perform return-to-center actions with clear retreat directions in advance while retaining pitch, roll, and yaw transient motion cues. This avoids irrelevant return-to-center components weakening effective attitude cues, directly reducing the probability of actuator travel saturation, attitude clipping, and reverse impact, and minimizing the inconsistency between the motion platform attitude and changes in the visual and instrument data. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0036] Figure 1 This is a flowchart of the method of the present invention.

[0037] Figure 2 This is a flowchart of the system modules of the present invention.

[0038] Figure 3 This is a flowchart illustrating the calculation of workspace constraint parameters for the motion platform of the present invention.

[0039] Figure 4This is a flowchart of the actuator saturation prediction and washing filter process of the present invention.

[0040] Figure 5 This is a flowchart of the phase advance compensation control for the wash-out return component of the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1, please refer to Figure 1 , Figure 3 , Figure 4 and Figure 5 As shown in this embodiment, a method for controlling the flight attitude of a flight simulator includes:

[0043] The simulated aircraft motion state signal output by the flight dynamics model is acquired, as well as the attitude feedback signal of the motion platform and the stroke feedback signal of each actuator; wherein, the simulated aircraft motion state signal includes pitch angle, roll angle, yaw angle, angular velocity and angular acceleration.

[0044] In this embodiment, the simulated aircraft motion state signal output by the flight dynamics model is first acquired, and the pose feedback signal of the motion platform and the stroke feedback signal of each actuator are acquired simultaneously. The flight dynamics model is used to calculate the simulated aircraft motion state in real time based on the current training subject, pilot control input, environmental disturbance parameters, and aircraft performance parameters.

[0045] Specifically, the pitch angle, roll angle, and yaw angle are used to characterize the attitude changes of the simulated aircraft relative to the geographic coordinate system or the body coordinate system; the angular velocity is used to characterize the rate of attitude change of the simulated aircraft; and the angular acceleration is used to characterize the trend of attitude change of the simulated aircraft. By simultaneously acquiring the attitude angle, angular velocity, and angular acceleration, complete attitude motion input can be provided for subsequent washing and filtering processing, avoiding the problems of response lag or insufficient motion cues that occur when control is based solely on attitude angle changes. The pose feedback signal of the motion platform is obtained from the attitude sensor, encoder, or platform controller installed on the motion platform. The pose feedback signal includes at least one of the current pitch attitude, roll attitude, yaw attitude, and platform center of mass position of the motion platform. The stroke feedback signal of each actuator is obtained from the displacement detection element installed on the electric cylinder, hydraulic cylinder, or servo cylinder, and is used to characterize the actual extension length or stroke position of each actuator at the current moment.

[0046] Based on the posture feedback signal and the stroke feedback signal of each actuator, the workspace constraint parameters of the motion platform are calculated. The workspace constraint parameters include the attitude angle limit margin and the remaining stroke of the actuator.

[0047] In this embodiment, the motion platform is a six-degree-of-freedom parallel motion platform, which includes multiple retractable actuators. The coordinates of the fixed-end hinge points, the coordinates of the moving-end hinge points, and the allowed minimum and maximum strokes of each actuator are pre-stored. Based on the pose feedback signal of the motion platform, the current position vector of the platform's moving coordinate system origin in the base coordinate system is obtained. And the rotation matrix determined by the current pitch angle, roll angle, and yaw angle. The current pose, represented by the current position vector and the rotation matrix, is determined as the reference pose.

[0048] To reduce the impact of platform structural parameter errors, hinge gaps, and sensor zero-position deviations on the attitude angle limit margin calculation results, inverse kinematics calculations are performed based on the reference pose to obtain the theoretical travel of each actuator in the current pose. The [further details are needed for a complete translation.] The difference between the stroke feedback value of the first actuator and the corresponding theoretical stroke is used to obtain the stroke feedback value of the second actuator. The stroke correction amount for each actuator. When calculating the virtual actuator stroke subsequently, the stroke correction amount is added to the corresponding theoretical virtual stroke.

[0049] The pitch, roll, and yaw axes are used as the attitude axes to be searched. For any attitude axis to be searched, keeping the platform's current position and the other two attitude angles unchanged, a preset virtual angle increment is successively added in the positive direction to the current attitude angle of the attitude axis to obtain multiple positive virtual attitudes; and the preset virtual angle increment is successively added in the negative direction to obtain multiple negative virtual attitudes. For example, the first... The virtual attitude angle corresponding to this search is: ;in, The current attitude angle of the attitude axis to be searched. To preset the virtual angle increment, This is the direction indicator; 1 represents a positive direction and -1 represents a negative direction.

[0050] Inverse kinematics calculations are performed based on the rotation matrices corresponding to each virtual pose. The actuator in the first The theoretical virtual journey under a virtual posture is: ;in, For the first The coordinates of the hinge point of the actuator's moving end in the platform's moving coordinate system. The coordinates of its fixed-end hinge point in the base coordinate system are given. The theoretical virtual stroke is added to the corresponding stroke correction amount to obtain the corrected virtual actuator stroke.

[0051] Each corrected virtual actuator stroke is compared with the minimum and maximum stroke boundaries of the corresponding actuator. When all virtual actuator strokes are within the corresponding stroke boundaries, a preset virtual angle increment is added along the current direction. When at least one virtual actuator stroke reaches or crosses the corresponding stroke boundary for the first time, the virtual attitude is determined as the boundary virtual attitude, and the actuator number and boundary type that reached the stroke boundary are recorded.

[0052] Optionally, to improve the calculation accuracy of the boundary attitude angle, the boundary virtual attitude and the previous virtual attitude that has not reached the travel boundary are combined to form an angle search interval. A binary search is performed within the angle search interval. Each time, the virtual travel of each actuator is recalculated based on the virtual attitude corresponding to the midpoint of the interval, and the search interval is narrowed according to whether the virtual travel has reached the travel boundary, until the angle width of the search interval is less than the preset angle accuracy. The attitude angle on the side that has not crossed the travel boundary is determined as the attitude limit angle in that direction.

[0053] The positive attitude angle limit margin is obtained by subtracting the corresponding current attitude angle from the attitude limit angle obtained in the positive direction for each attitude axis to be searched; the negative attitude angle limit margin is obtained by subtracting the corresponding current attitude angle from the attitude limit angle obtained in the negative direction. .

[0054] This yields attitude angle limit margins for pitch, roll, and yaw angles in both positive and negative directions. These attitude angle limit margins are used to predict whether the motion platform will experience actuator saturation risk when executing target motion commands.

[0055] In this embodiment, to accurately obtain the remaining available travel of each actuator in the current motion state, the current spatial pose of the motion platform is first determined based on the pose feedback signal of the motion platform. Specifically, based on the platform position parameters and attitude angle parameters in the pose feedback signal, the current coordinate system of the motion platform is established, and inverse kinematics calculations are performed in conjunction with the structural parameters of the motion platform to obtain the theoretical travel value corresponding to each actuator in the current theoretical pose.

[0056] Due to factors such as actuator installation errors, transmission backlash, and zero-point drift of the displacement detection element during actual operation, there is a certain deviation between the theoretical stroke and the actual feedback stroke of each actuator. Therefore, the theoretical stroke corresponding to the i-th actuator is... The actual trip in the trip feedback signal Perform difference calculation to obtain the stroke deviation of the i-th actuator: ;in, This represents the actual position correction amount of the i-th actuator due to installation errors and feedback errors.

[0057] The preset stroke boundary is dynamically corrected based on the stroke deviation. Specifically, the maximum and minimum allowable strokes for each actuator are pre-stored, and the stroke deviations are superimposed onto the preset stroke boundary in the corresponding directions to obtain an equivalent stroke boundary that matches the current motion state. Specifically, for actuators moving in the extension direction, the stroke deviation is superimposed onto the maximum stroke boundary; for actuators moving in the contraction direction, the stroke deviation is superimposed onto the minimum stroke boundary to eliminate the deviation between the theoretical model boundary and the actual execution boundary.

[0058] Furthermore, the attitude change trend of the motion platform within the current control cycle is determined based on the simulated aircraft motion state signal, and the extension / retraction direction corresponding to each actuator is calculated based on the attitude change trend. Specifically, the angular velocity and angular acceleration in the simulated aircraft motion state signal are input into the actuator motion mapping model, and the extension or retraction state corresponding to each actuator in the next control cycle is determined based on the spatial arrangement relationship of each actuator in the motion platform.

[0059] After determining the extension and retraction directions of each actuator, the corresponding equivalent stroke boundary is selected according to the extension and retraction directions, and the difference between the actual stroke feedback signal of each actuator and the equivalent stroke boundary is calculated. When the actuator is moving in the extension direction, the difference between the equivalent maximum stroke boundary and the current actual stroke is calculated to obtain the remaining extension stroke of the actuator; when the actuator is moving in the retraction direction, the difference between the current actual stroke and the equivalent minimum stroke boundary is calculated to obtain the remaining retraction stroke of the actuator.

[0060] The above method enables dynamic correction of the actuator's theoretical boundary based on the current actual state of the motion platform, and determines the effective remaining stroke in conjunction with the next motion trend. This ensures that the obtained remaining stroke of the actuator reflects the actual usable motion space of the motion platform within the current control cycle, providing an accurate data basis for subsequent judgment of actuator saturation risk and adjustment of motion platform control commands.

[0061] Based on the simulated aircraft motion state signal and workspace constraint parameters, rolling prediction is performed within a preset prediction time domain to obtain the actuator saturation prediction point. Based on the actuator saturation prediction point, the simulated aircraft motion state signal is subjected to washout filtering to obtain high-frequency motion clue components and washout return components.

[0062] In this embodiment, the controller acquires the simulated aircraft motion state signal and workspace constraint parameters at fixed control cycles, and updates the prediction result once in each control cycle. Let the current control time be... The control cycle is The preset prediction time domain includes If there are 1 prediction step, then the prediction time domain length is 1. The simulated aircraft motion state signal includes an attitude angle vector composed of the current pitch angle, roll angle, and yaw angle. Angular velocity vector and angular acceleration vector .

[0063] Within the preset prediction time domain, taking the current simulated aircraft motion state as the initial state, the following formula is used to recursively deduce the... Predicted angular velocity and predicted attitude angle at each prediction time: , ;in, When a new angular acceleration is obtained in an adjacent control cycle, the initial state of the previous control cycle is replaced with the new simulated aircraft motion state, and subsequent cycles are recalculated. A number of prediction moments are generated, thus forming a rolling forecast.

[0064] Based on the current attitude angles of the motion platform and the attitude angle limit margins in each direction, allowable attitude ranges are established for pitch, roll, and yaw directions. Taking the pitch direction as an example, the current pitch angle is... The positive attitude angle limit margin and the negative attitude angle limit margin are respectively and Then the allowable range of pitch angle is: The roll direction and yaw direction are determined in the same way to determine the corresponding allowable attitude range. The predicted attitude angle at each prediction time is compared with the corresponding allowable attitude range, and actuator stroke calculation is performed only for the predicted attitude within the allowable attitude range; when the predicted attitude angle reaches the boundary of the allowable attitude range, the predicted attitude is taken as the boundary attitude of the motion platform's executable space.

[0065] For the At each predicted moment, a predicted rotation matrix is ​​established based on the predicted attitude angle, while keeping the current platform center of mass position unchanged. The predicted stroke of each actuator is calculated using the inverse kinematics model of the motion platform. The actuator in the first The predicted journey at each predicted time point is represented as follows: ;in, Let the vector be the current centroid position of the motion platform. For the first The prediction rotation matrix corresponding to each prediction time. For the first The coordinates of the hinge point of the actuator's moving end in the platform's moving coordinate system. This corresponds to the coordinates of the fixed-end hinge point in the base coordinate system.

[0066] Predicting the trip With the Current travel feedback value of each actuator By subtracting the values, we obtain the predicted change in travel distance: ;when When it is greater than zero, determine the first The actuator is positioned in the predicted extension direction, and the remaining extension stroke of the actuator is selected. As the available range of motion; when When less than zero, determine the first Each actuator is positioned in the predicted contraction direction, and the remaining contraction stroke of that actuator is selected. As the available range of motion. The actuator in the first The trip occupancy rate at each predicted time point is calculated using the following formula: The stroke occupancy rate is compared sequentially at each predicted time, from most recent to furthest. When the stroke occupancy rate of any actuator first reaches the preset saturation value... When the time is determined, the corresponding prediction time is set as the actuator saturation prediction point. The actuator saturation prediction point includes the prediction step number that reaches the preset saturation value. The saturation actuator number and the predicted stroke of the actuator are specified. The preset saturation value is less than or equal to 1; in one specific embodiment, the preset saturation value is set to 0.85 to retain control margin before the actuator reaches the actual stroke boundary.

[0067] Calculate the saturation time distance based on the actuator saturation prediction point: And based on the predicted stroke of the saturation actuator at the predicted saturation point of the actuator, calculate the predicted remaining stroke: ;in, This represents the current remaining stroke of the saturation actuator along the predicted extension / retraction direction. This represents the predicted stroke change when the actuator reaches the actuator saturation prediction point.

[0068] The cutoff adjustment coefficient for the wash-out filter is determined based on the predicted remaining travel distance and saturation time distance. The cutoff adjustment coefficient is calculated according to the following formula: ;in, and For pre-defined non-negative weights, and ; To set the upper limit of the cutoff adjustment coefficient. When there is no actuator saturation prediction point within the preset prediction time domain, let... Therefore, the smaller the predicted remaining stroke and the closer the actuator saturation prediction point is to the current control moment, the larger the cutoff adjustment coefficient.

[0069] For the pitch, roll, and yaw attitude directions, the equivalent attitude input for the current control cycle is constructed based on the attitude angle, angular velocity, and angular acceleration, respectively: ;in, These represent the pitch, roll, and yaw directions, respectively. The fundamental cutoff frequency... Multiplying this by the cutoff adjustment coefficient yields the actual cutoff frequency for the current control cycle: ;

[0070] Dynamic frequency decomposition is performed on the equivalent attitude input in each direction using the actual cutoff frequency. Taking a first-order discrete high-pass filter as an example, the filter coefficients are: ;No. The high-frequency motion cue components in each attitude direction are calculated according to the following formula: ;in, This represents the high-frequency motion cue component from the previous control cycle. Subtracting the equivalent attitude input from the high-frequency motion cue component yields the low-frequency attitude change: The low-frequency attitude change is reversed to obtain the washout return component: .

[0071] Composed of pitch, roll and yaw The high-frequency motion cue components are composed of the corresponding The washout and centering components are composed of a high-frequency motion cue component and a high-frequency motion cue component. The high-frequency motion cue component is used to retain transient motion cues generated by rapid attitude changes of the simulated aircraft, while the washout and centering component is used to drive the motion platform to release the occupied attitude angle space and actuator stroke. The washout and centering component serves as the input for subsequent phase lead compensation and is synthesized with the high-frequency motion cue component after phase lead compensation is completed to form the desired pose command of the motion platform.

[0072] Phase advance compensation is performed on the wash-out return component, and it is synthesized with the high-frequency motion clue component to obtain the desired pose command of the motion platform. The target stroke of each actuator is solved according to the desired pose command of the motion platform, and the action of each actuator is controlled.

[0073] In this embodiment, the motion platform adopts a parallel motion platform including multiple retractable actuators. Let the current control cycle be the [number missing]. There are [number] control cycles, with a control cycle length of [length]. The high-frequency motion clue component and the washed-back component obtained from the aforementioned washing-out filtering process are respectively expressed as: , ;in, , and These are the high-frequency motion cue components in the pitch, roll, and yaw directions, respectively. , and These represent the wash-out and return components in the corresponding directions; superscript This represents the matrix transpose. In this embodiment, the high-frequency motion cue component and the washout return component are both angular increments relative to the center attitude of the motion platform, and they use the same coordinate system, rotation direction, and angular unit.

[0074] The actuator saturation prediction point includes the saturation prediction step number. Saturation actuator number The predicted extension / retraction direction and corresponding predicted attitude of the saturated actuator. The time distance from the current control moment to the actuator's saturation prediction point is: ;in, For saturation time distance, This is the prediction step number corresponding to when the actuator first reaches the preset saturation value. When more than two actuators reach the preset saturation value within the same prediction step, the actuator with the highest stroke occupancy rate is identified as the saturated actuator.

[0075] The predicted stretching direction of the saturation actuator is represented as follows: When the saturation actuator extends toward the equivalent maximum stroke boundary, let When the saturation actuator contracts toward the equivalent minimum stroke boundary, let Therefore, with The opposite direction of actuator stroke change is the direction that moves the saturation actuator away from the corresponding equivalent stroke boundary.

[0076] To determine the portion of the washout return component capable of driving the saturated actuator away from the corresponding equivalent travel boundary, the travel sensitivity of the saturated actuator relative to pitch, roll, and yaw angles is calculated using the predicted attitude corresponding to the actuator's saturation prediction point as the calculation benchmark. For any attitude direction... The stroke sensitivity is calculated according to the following formula: ;in, For saturation actuators to the first Travel sensitivity in each attitude direction; These represent the pitch direction, roll direction, and yaw direction, respectively. This is the predicted attitude angle vector corresponding to the actuator saturation prediction point; For the first Preset angular disturbance amount in each attitude direction; For the first Unit vectors corresponding to each attitude direction; This represents the saturated actuator stroke calculated using the inverse kinematics model of the motion platform. When calculating the actuator stroke corresponding to the two perturbed attitudes, the position components of the motion platform are maintained as the position components corresponding to the actuator saturation prediction points.

[0077] Based on the aforementioned stroke sensitivity, calculate the boundary approximation of the wash-out return component to the saturation actuator in each attitude direction: ;in, For the first The boundary approximation quantity corresponding to each attitude direction.

[0078] when When this occurs, it indicates that the washout return component in this attitude direction causes the saturation actuator to produce a stroke change opposite to the predicted extension direction, and this washout return component is determined as the target return component.

[0079] when At that time, the washout return component in that direction is not determined as the target return component.

[0080] The A preset threshold for judging stroke changes is used to eliminate misjudgments of direction caused by errors in inverse kinematics calculations.

[0081] Based on the above judgment results, establish a selection matrix: ;in, Represents a diagonal matrix; when the corresponding attitude direction satisfies season Otherwise The target return component is: Therefore, the target centering component retains only the pitch, roll, or yaw centering components that can move the saturation actuator away from the equivalent travel boundary.

[0082] The phase lead is determined based on the saturation time distance. Let the preset prediction time domain length be... ,in, This represents the number of prediction steps included in the time domain. First, calculate the saturation pressure coefficient: ;in, The saturation pressure coefficient; Indicates will Limited to the range The closer the actuator saturation prediction point is to the current control moment, the better. The smaller the value, the greater the saturation pressure coefficient. The larger.

[0083] The phase lead compensation coefficient is determined based on the saturation urgency coefficient: ;in, and The predefined compensation coefficient boundary is set, and it satisfies: Further determine the time constants of the leading and lagging components: , ;in, The time constant of the lead element, The time constant of the lag element; and These are the lower and upper limits of the time constant of the lead element, respectively. Because... The corresponding phase lead element is: ;in, Let be the Laplace transform variable. The maximum phase lead that this phase lead element can provide is: The closer the actuator saturation prediction point is to the current control moment, The smaller the value, the greater the maximum phase lead. The larger the value, the earlier the target return component that can release the actuator stroke is output.

[0084] The phase lead element is converted into a discrete form using a bilinear transformation. For any selected target return direction, the advance output result is calculated according to the following formula: ;in, For the first The output results of each control cycle are delivered in advance; This is an advance output of the results from the previous control cycle; and These represent the target homing components for the current control cycle and the previous control cycle, respectively. The discrete compensation coefficients are as follows: , , This discrete phase lead compensation only uses data from the current control cycle and historical control cycles, without needing to obtain future actual sample values, and can be executed in real time by the motion platform controller.

[0085] The target return-center component after phase advance compensation is combined with the unselected washed-out return-center component to obtain the complete compensated return-center component: ;in, This is a vector composed of the advance output results of the centering direction of each target; It is a third-order identity matrix. Therefore, only the return-center component that can move the saturation actuator away from the corresponding equivalent stroke boundary is phase-advanced compensated, while the other return-center components maintain their original output.

[0086] The compensated midpoint component is synthesized with the high-frequency motion cue component. Let the preset midpoint attitude of the motion platform be... The desired attitude angle without limit treatment is: ;in, This is the desired attitude angle vector without any limiting processing. Based on the aforementioned attitude angle limiting margin, the desired attitude angle is constrained. For the first... For each attitude direction, the constrained desired attitude angle is: ;in, The current feedback attitude angle of the motion platform; and These are the negative attitude angle limit margin and the positive attitude angle limit margin, respectively, for the corresponding attitude directions.

[0087] This embodiment mainly controls the flight attitude of the motion platform. Therefore, the position component in the desired pose command of the motion platform is set as a preset median position vector. The attitude components are set as the constrained desired attitude angle vectors. The desired pose command for the motion platform is obtained: The desired rotation matrix is ​​constructed based on the pitch, roll, and yaw angles. The rotation sequence is pre-set as roll first, then pitch, then yaw; therefore, the desired rotation matrix is: ;in, , and These are the rotation matrices around the roll axis, pitch axis, and yaw axis, respectively.

[0088] The inverse kinematics solution is performed for the desired pose command of the motion platform. The theoretical target stroke of each actuator is: ;in, For the first The theoretical target stroke of an actuator; For the first The coordinates of the hinge point of the actuator's moving end in the platform's moving coordinate system; The coordinates of the fixed-end hinge point in the base coordinate system; symbol This represents the Euclidean length of the vector.

[0089] By limiting the theoretical target stroke within the corresponding equivalent stroke boundary, the actual target stroke of the actuator is obtained: ;in, and The first The equivalent minimum travel boundary and the equivalent maximum travel boundary of each actuator.

[0090] The stroke tracking error is obtained by subtracting the actuator's target stroke from the corresponding stroke feedback signal. ;in, For the first The stroke feedback value of each actuator in the current control cycle. When At that time, control the first The actuator extends; when At that time, control the first The actuator contracts.

[0091] In one specific control method, a proportional-integral controller is used to generate actuator speed commands: , ;in, For the first The integral quantity of the error of an actuator; For proportional gain; For integral gain; For actuator speed command; This represents the maximum permissible speed of the actuator. Based on the speed command, the actuator position command for the next control cycle is generated. The motion platform controller sends the actuator position commands to the corresponding servo drives, causing each actuator to track its target stroke. When the absolute value of the stroke tracking error is less than a preset position error threshold, the speed command of the corresponding actuator is set to zero to prevent the actuator from reciprocating near the target position.

[0092] This invention uses pitch, roll, yaw, angular velocity, and angular acceleration output from a flight dynamics model as target motion inputs. Simultaneously, it acquires the attitude of the motion platform and the stroke of each actuator, calculating the attitude angle limit margin and the remaining actuator stroke. Then, within a preset prediction time domain, it performs rolling predictions of future attitude and actuator stroke, pre-determining actuator saturation prediction points that may reach stroke boundaries. Based on these prediction points, it dynamically adjusts the washout filtering process, decomposing the simulated aircraft motion state signal into high-frequency motion cue components and washout return-to-center components. Next, it selects the return-to-center component that can move the saturated actuators away from the equivalent stroke boundary, and performs phase advance compensation based on the time distance between the saturation prediction point and the current moment, allowing the platform to release occupied motion space in advance. Finally, it synthesizes the compensated washout return-to-center component with the high-frequency motion cue components to generate the desired attitude command for the motion platform. The target stroke of each actuator is obtained through inverse kinematics, and the actuator actions are controlled based on the deviation between the target stroke and the feedback stroke. This approach retains transient flight attitude perception while reducing actuator saturation, attitude clipping, and reverse impact.

[0093] Example 2, please refer to Figure 2 As shown in this embodiment, a control system for the flight attitude of a flight simulator includes:

[0094] The signal acquisition module acquires the simulated aircraft motion state signals output by the flight dynamics model, and acquires the attitude feedback signals of the motion platform and the stroke feedback signals of each actuator; wherein, the simulated aircraft motion state signals include pitch angle, roll angle, yaw angle, angular velocity and angular acceleration;

[0095] The workspace constraint parameter calculation module calculates the workspace constraint parameters of the motion platform based on the pose feedback signal and the stroke feedback signal of each actuator. The workspace constraint parameters include the attitude angle limit margin and the remaining stroke of the actuator.

[0096] The saturation prediction and washout filtering module performs rolling prediction within a preset prediction time domain based on the simulated aircraft motion state signal and workspace constraint parameters to obtain the actuator saturation prediction point, and performs washout filtering on the simulated aircraft motion state signal based on the actuator saturation prediction point to obtain the high-frequency motion clue component and the washout return component.

[0097] The actuator control module performs phase advance compensation on the wash-out return component, synthesizes it with the high-frequency motion clue component to obtain the desired pose command of the motion platform, solves the target stroke of each actuator according to the desired pose command of the motion platform, and controls the action of each actuator.

[0098] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for controlling the flight attitude of a flight simulator, characterized in that, include: The simulated aircraft motion state signal output by the flight dynamics model is acquired, as well as the attitude feedback signal of the motion platform and the stroke feedback signal of each actuator; wherein, the simulated aircraft motion state signal includes pitch angle, roll angle, yaw angle, angular velocity and angular acceleration; Based on the posture feedback signal and the stroke feedback signal of each actuator, the workspace constraint parameters of the motion platform are calculated. The workspace constraint parameters include the posture angle limit margin and the remaining stroke of the actuator. Based on the simulated aircraft motion state signal and workspace constraint parameters, rolling prediction is performed within a preset prediction time domain to obtain the actuator saturation prediction point. Based on the actuator saturation prediction point, the simulated aircraft motion state signal is subjected to washout filtering to obtain high-frequency motion clue components and washout return components. Phase advance compensation is performed on the wash-out return component, and it is synthesized with the high-frequency motion clue component to obtain the desired pose command of the motion platform. The target stroke of each actuator is solved according to the desired pose command of the motion platform, and the action of each actuator is controlled.

2. The method for controlling the flight attitude of a flight simulator according to claim 1, characterized in that, The calculation of attitude angle limit margin includes: The current pose represented by the pose feedback signal is determined as the reference pose. Preset virtual angle increments are successively superimposed on the reference pose along the positive and negative directions of pitch, roll and yaw to obtain the corresponding virtual attitude sequence. The inverse kinematics solution is performed on each virtual pose in the virtual pose sequence, and the stroke correction value obtained by subtracting the theoretical stroke under the current pose from the stroke feedback value is added to the corresponding theoretical virtual stroke to obtain the corrected virtual stroke; When any modified virtual stroke first reaches the preset stroke boundary of the corresponding actuator, the difference between the corresponding virtual attitude angle and the current attitude angle is determined as the attitude angle limit margin in the corresponding direction.

3. The method for controlling the flight attitude of a flight simulator according to claim 2, characterized in that, The calculation of attitude angle limit margin also includes: The virtual attitude at which the corrected virtual travel first reaches the preset travel boundary is combined with the previous virtual attitude that did not reach the preset travel boundary to form an angle search interval; inverse kinematics is solved using the midpoint attitude of the angle search interval, and the sub-interval containing the travel boundary is retained based on whether the obtained corrected virtual travel reaches the preset travel boundary; the angle search interval is repeatedly reduced until the interval width is no greater than the preset angle accuracy, and the attitude angle on the side that did not reach the preset travel boundary is determined as the attitude limit angle.

4. The method for controlling the flight attitude of a flight simulator according to claim 2, characterized in that, The calculation of the remaining stroke of the actuator includes: Based on the pose feedback signal, the motion platform is solved by inverse kinematics to obtain the theoretical stroke of each actuator in the current pose. The corresponding stroke feedback value is then subtracted from the theoretical stroke to obtain the stroke correction amount of each actuator. The stroke correction amount is superimposed on the preset minimum stroke boundary and preset maximum stroke boundary of the corresponding actuator to obtain the equivalent minimum stroke boundary and equivalent maximum stroke boundary; The extension and retraction direction of each actuator is determined based on the simulated aircraft motion state signal, and the remaining stroke of each actuator is obtained by subtracting the stroke feedback value from the corresponding equivalent stroke boundary according to the extension and retraction direction.

5. The method for controlling the flight attitude of a flight simulator according to claim 4, characterized in that, Rolling prediction is performed within a preset prediction time domain to obtain the actuator saturation prediction point, including: Using the current attitude angle, angular velocity, and angular acceleration as the initial state, the attitude change sequence in the preset prediction time domain is obtained by recursion according to the control cycle, and the attitude change sequence is superimposed on the current pose of the motion platform to obtain the predicted attitude sequence. Inverse kinematics is performed on each predicted posture to obtain the predicted stroke of each actuator. The corresponding remaining stroke is selected according to the predicted extension and retraction direction, and the stroke occupancy rate of the predicted stroke change to the remaining stroke is calculated. Based on the prediction time from near to far, the prediction time when the stroke occupancy rate of any actuator first reaches the preset saturation value, the actuator number, and the predicted extension / retraction direction are determined as the actuator saturation prediction point.

6. The method for controlling the flight attitude of a flight simulator according to claim 5, characterized in that, The washing and filtering process includes: Based on the predicted stroke change and remaining stroke corresponding to the actuator saturation prediction point, the predicted remaining stroke is obtained, and the saturation time distance is obtained based on the predicted time of the actuator saturation prediction point. Based on the ratio of the predicted remaining stroke to the remaining stroke and the ratio of the saturation time distance to the preset prediction time domain length, the cutoff adjustment coefficient is determined, and the preset base cutoff frequency is adjusted using the cutoff adjustment coefficient. The equivalent attitude input formed by attitude angle, angular velocity and angular acceleration is high-pass filtered according to the adjusted cutoff frequency to obtain the high-frequency motion clue component. The difference between the equivalent attitude input and the high-frequency motion clue component is reversed to obtain the washout return component.

7. The method for controlling the flight attitude of a flight simulator according to claim 6, characterized in that, Before performing phase lead compensation on the washed-out return component, the following is also included: Based on the predicted attitude corresponding to the saturation prediction point of the actuator, positive and negative angle perturbations are applied along the pitch, roll and yaw directions respectively, and the inverse kinematics of the perturbed attitude is solved to obtain the stroke sensitivity of the saturation actuator relative to each attitude direction. By associating the stroke sensitivity of each attitude direction, the corresponding washout return component, and the predicted extension / retraction direction of the saturation actuator with signs, the boundary approach amount corresponding to each washout return component is obtained. The washout return component with a negative boundary approach value and an absolute value greater than a preset judgment threshold is determined as the target return component that drives the saturation actuator away from the corresponding equivalent stroke boundary.

8. The method for controlling the flight attitude of a flight simulator according to claim 7, characterized in that, Phase lead compensation includes: The saturation urgency coefficient is determined based on the ratio of the saturation time distance to the preset prediction time domain length, and the lead time constant and lag time constant of the phase lead element are determined according to the saturation urgency coefficient, so that the lag time constant is smaller than the lead time constant. The phase lead element, composed of the lead time constant and the lag time constant, is discretized according to the control cycle. The discretized phase lead element is then used to filter the target return-to-center component to obtain the target return-to-center component that is output in advance. The pre-output target return component is combined with the original wash-out return component that was not identified as the target return component to obtain the compensated wash-out return component.

9. The method for controlling the flight attitude of a flight simulator according to claim 8, characterized in that, Obtain the desired pose command from the motion platform and control the actions of each actuator, including: The compensated washout return component and the high-frequency motion clue component are superimposed on the preset neutral attitude of the motion platform according to the pitch, roll and yaw directions, and the superposition result is restricted by the attitude angle limit margin in each direction to obtain the desired pose command of the motion platform. The desired pose command of the motion platform is solved by inverse kinematics, and the obtained theoretical target stroke is restricted between the equivalent minimum stroke boundary and the equivalent maximum stroke boundary of the corresponding actuator to obtain the target stroke of each actuator; the target stroke of each actuator is subtracted from the corresponding stroke feedback value, and the actuator extension and retraction control command is generated based on the obtained stroke tracking error to control each actuator to track the target stroke.

10. A control system for the flight attitude of a flight simulator, used to implement the control method for the flight attitude of a flight simulator according to any one of claims 1-9, characterized in that, It includes: a signal acquisition module, a workspace constraint parameter calculation module, a saturation prediction and washout filtering module, and an actuator control module.

Images