A multi-rotor unmanned aerial vehicle cooperative flight control method based on a plunger pump motor

Abstract

The application discloses a kind of multi-rotor unmanned aerial vehicle cooperative flight control methods based on plunger pump motor, specifically relates to unmanned aerial vehicle flight control field, including obtaining the relative pose measurement data and inertial measurement data of formation unmanned aerial vehicle, and according to formation task requirement and safety constraint, the lift reference instruction of each unmanned aerial vehicle in current control period is calculated;The application is by in short time window, the identification sequence of each unmanned aerial vehicle is injected into lift channel with amplitude limiting, and is matched with window inertia response alignment to estimate online washout coupling coefficient, and then based on the washout coupling coefficient generation and execution cooperative compensation lift instruction is carried out to implement cooperative process control, to solve in near formation flight, washout coupling relationship fast time-varying and cannot be pre-modeled, leading to existing cooperative control cannot identify coupling strength online, thereby generating formation error bias and inducing cooperative oscillation problem.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) flight control technology, and more specifically, to a collaborative flight control method for multi-rotor UAVs based on a plunger pump motor. Background Technology

[0002] In the field of multi-rotor UAV collaborative flight control, the mainstream practices in the industry mainly revolve around keeping multiple UAVs in formation and suppressing relative deviations under the same mission objective. Typically, the lift and attitude control of each UAV are treated as independent controllable objects. At the formation level, consistency control, lead and follow or model-based formation maintenance strategies are adopted, supplemented by empirical downwash compensation parameters or slow adaptive correction. The updates of collaborative control quantities are constrained by process control methods, thereby achieving stable collaboration without significantly increasing system complexity. However, in real close-range escort or narrow passage scenarios, multiple multi-rotor aircraft driven by plunger pump motors often need to maintain a small-space formation within a limited space and meet hard constraints such as limited relative position error, limited attitude jitter, limited cooperative control update cycle, and communication jitter. At this time, the downwash airflow of one aircraft will significantly change the rotor inflow conditions of the adjacent aircraft, causing observable lift deviation and attitude perturbation of the adjacent aircraft under the same thrust command. Moreover, the coupling strength will change rapidly with relative distance, relative speed, and external wind field in a short period of time, causing the mainstream approach to consistently expose a bottleneck: the formation error exhibits a systematic bias and transforms into periodic oscillations in some maneuvers. Specific verifiable phenomena include repeated opening and closing of the formation in terms of altitude and spacing fluctuations even without significant external disturbances, and the compensation command of individual aircraft continuously increasing but the error not converging. The root cause is that the cooperative control assumes that the thrust of each aircraft is independent and responsible for itself, but it cannot identify who is affecting whom and how much the time-varying coupling relationship is at this moment during operation and implement effective process control of the coupling relationship, thus causing systematic deviations in the direction and magnitude of compensation. Therefore, under the constraints of close-range cooperative flight and the time-varying downwash coupling that cannot be pre-modeled, how to determine the downwash coupling relationship between multiple aircraft in an online identifiable manner and generate stable cooperative control quantities accordingly to avoid formation error bias and cooperative oscillation has become a pressing technical problem that needs to be solved. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a cooperative flight control method for multi-rotor UAVs based on a plunger pump motor. This method estimates the downwash coupling coefficient online by injecting the identification sequence of each UAV into the lift channel with limited amplitude within a short time window and aligning it with the inertial response within the window. Then, based on the downwash coupling coefficient, a cooperative compensation lift command is generated and executed to implement cooperative process control, thereby solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a cooperative flight control method for a multi-rotor unmanned aerial vehicle based on a plunger pump motor, comprising: S1. Acquire the relative pose measurement data and inertial measurement data of the formation drones, and calculate the lift reference command of each drone in the current control cycle according to the formation mission requirements and safety constraints. S2. For each UAV in the formation, collect the plunger pump motor pressure sampling signal of the UAV and determine the starting point of the short time window from the pressure sampling signal. Generate the identification increment of the identification sequence corresponding to the UAV according to the starting point and limit the identification increment according to the current lift margin of the UAV to obtain the limited identification increment. S3. The lift reference command of each UAV is superimposed with the corresponding limited recognition increment to obtain the recognition lift command, and the recognition lift command is input into the plunger pump motor controller to drive the rotor to generate thrust, thereby obtaining the inertial measurement response data within a short time window. S4. For each UAV, the inertial measurement response data is aligned within the window starting point to obtain an aligned response sequence. The aligned response sequence is then matched and statistically calculated with the identification sequences corresponding to other UAVs in the formation to obtain a set of downwash coupling coefficients.

[0005] In a preferred embodiment, S5, calculate the collaborative compensation increment based on the set of downwash coupling coefficients and the lift reference command, and superimpose the collaborative compensation increment onto the lift reference command to obtain the collaborative lift command; S6. Input the coordinated lift command of each UAV into the corresponding plunger pump motor controller and execute it in the next control cycle. At the same time, update the number of cycles in the window and identify the incremental limit rule according to the set of downwash coupling coefficients and the formation error sequence to obtain the updated parameters.

[0006] In a preferred embodiment, S1 includes: S1-1. Calculate the vertical height deviation of the formation based on the relative pose measurement data, divide the height deviation by the current control cycle duration to obtain the target vertical velocity, and limit the target vertical velocity within the speed range allowed by the safety constraints, and output the target vertical velocity. S1-2. Calculate the velocity difference based on the target vertical velocity and the actual vertical velocity in the inertial measurement data, divide the velocity difference by the current control cycle duration to obtain the target vertical acceleration, and limit the target vertical acceleration within the acceleration range allowed by the safety constraints, and output the target vertical acceleration. S1-3. Calculate the lift correction amount based on the lift reference command of the previous control cycle, the change in vertical acceleration in the inertial measurement data, and the lift margin of each UAV. Then, superimpose the lift correction amount onto the lift reference command of the previous control cycle and perform trimming and margin redistribution according to the lift margin to obtain the lift reference command of the current control cycle, and output the lift reference command.

[0007] In a preferred embodiment, S2 includes: S2-1. Collect the plunger pump motor pressure sampling signal for each UAV, calculate the average pressure, ripple amplitude and period difference within the sliding window, and write the pressure valid mark and period duration when the period difference is less than the allowable difference and the ripple amplitude is not zero, and output the pressure valid mark and period duration. S2-2. When the pressure valid flag is valid, determine the starting point of the short time window based on the adjacent peak times of the pressure sampling signal and write the starting point of the short time window. When the pressure valid flag is invalid, use the starting point of the short time window of the previous control cycle and write the back-off flag. Output the starting point of the short time window and the back-off flag. S2-3. Read the lift margin of the UAV, convert the lift margin into the upper limit of recognition increment according to the limit rule, and output the upper limit of recognition increment. S2-4. Read the drone's identifier and determine the corresponding identification sequence. Based on the starting point of the short time window, map the identification sequence to the identification increment. Then, trim the identification increment according to the upper limit of the identification increment to obtain the limited identification increment. Output the limited identification increment. S2-5. Calculate the vertical jitter amplitude based on the formation error sequence and inertial measurement response data, and perform gating judgment on the vertical jitter amplitude according to the amplitude limiting rule. When the gating fails, set the amplitude-limited recognition increment to zero and write a disabled flag. When the gating passes, write an enabled flag and retain the amplitude-limited recognition increment. Output the gating flag and the amplitude-limited recognition increment. S2-6. Calculate the matching value based on the recognition sequence and inertial measurement response data of the previous short time window, and verify the matching value according to the amplitude limiting rule. If the verification fails, write a re-examination flag and update the upper limit of the recognition increment according to the amplitude limiting rule. Then, trim the recognition increment again and perform gating to update the amplitude-limited recognition increment. If the verification passes, write a pass flag and output the amplitude-limited recognition increment.

[0008] In a preferred embodiment, S3 includes: S3-1. For each UAV, add the lift reference command to the recognition increment after limiting to obtain the recognition lift command. Divide the difference of the recognition lift command between adjacent control cycles by the control cycle duration to obtain the rate of change. Then, trim the rate of change according to the upper limit of the rate of change to obtain the speed-limited recognition lift command. S3-2. Input the speed-limited identification lift command into the plunger pump motor controller and execute it within a short time window. At the same time, collect inertial measurement response data within the short time window and align it with the starting point of the short time window to form a response sequence within the window. S3-3. Calculate the mean vertical acceleration and the vertical acceleration fluctuation for the response sequence within the window, and write the mean vertical acceleration and the vertical acceleration fluctuation into the inertial measurement response number.

[0009] In a preferred embodiment, S4 includes: S4-1. For each UAV, inertial measurement response data is extracted at the starting point of a short time window and arranged in the order of sampling time to obtain an aligned response sequence; S4-2. Calculate the matching value between the alignment response sequence and the corresponding identification sequence of each of the other UAVs in the formation, and normalize the matching value according to the amplitude of the identification sequence to obtain the set of downwash coupling coefficients. S4-3. Calculate the reconstructed response based on the set of downwashed coupling coefficients and the aligned response sequence, and obtain the residual by subtracting from the aligned response sequence. When the residual exceeds the allowable range, shrink the set of downwashed coupling coefficients proportionally and write a verification mark. When the residual does not exceed the allowable range, write a pass mark and output the set of downwashed coupling coefficients.

[0010] In a preferred embodiment, S5 includes: S5-1. For each UAV, multiply the lift reference command of other UAVs in the formation by the corresponding downwash coupling coefficient and sum them to obtain the downwash influence amount. S5-2. For each drone, the negative value of the downwash effect is used to obtain the original compensation value, and the original compensation value is trimmed according to the upper limit of the compensation range to obtain the collaborative compensation increment. S5-3. For each UAV, the collaborative compensation increment is superimposed on the lift reference command to obtain the collaborative lift command and output it.

[0011] In a preferred embodiment, S6 includes: S6-1. Input the coordinated lift command into the plunger pump motor controller and execute it in the next control cycle. At the same time, record the set of downwash coupling coefficients and the formation error sequence of this control cycle to obtain the record of this cycle. S6-2. Calculate the change in the set of washing coupling coefficients based on the current cycle record and the previous control cycle record, and calculate the error fluctuation based on the current cycle formation error sequence. Update the number of cycles within the window based on the change and fluctuation, and calculate the duration within the window based on the cycle duration and the number of cycles within the window. S6-3. Based on the number of cycles within the window, the amount of change, and the lift margin, update and identify the incremental limiting rules and form update parameters, then output the update parameters.

[0012] The technical effects and advantages of this invention are as follows: 1. By identifying the sequence injection and performing matching statistics based on the in-window aligned response, the washing coupling coefficient is obtained online and compensation is calculated to form a cooperative process control, which relatively reduces the formation error bias and cooperative oscillation; 2. By determining the starting point and cycle duration of the short time window from the peak pressure of the plunger pump, the responses of each machine have a unified alignment reference, which relatively reduces the coupling estimation deviation under communication jitter. 3. By converting the lift margin into an incremental upper limit and performing a trimming, and by setting the vertical jitter amplitude and formation error gating to zero, process control is implemented to relatively suppress the amplification of disturbances caused by the identification injection; 4. By reviewing the matching value of the previous short time window, if it is lower than the lower limit of the score, the upper limit of the recognition increment is lowered and the pruning and gating are re-implemented to complete the process control of the injection intensity, so that the injection intensity converges relatively with the recognizability. 5. By limiting the speed of the identified lift command according to the upper limit of the rate of change and collecting the alignment response sequence within a short time window, and by using the rate of change process control, the overshoot and attitude jitter caused by insufficient tracking of the plunger pump motor are relatively mitigated. 6. The coupling coefficient is checked by reconstructing the response residual. If it exceeds the limit, it is shrunk proportionally and rechecked. The number of cycles within the window and the limiting rules are updated according to the coefficient change and error fluctuation to improve stability under changing operating conditions. Attached Figure Description

[0013] Figure 1 This is a flowchart of the method steps of the present invention. Detailed Implementation

[0014] 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, and 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.

[0015] Refer to the instruction manual appendix Figure 1 The present invention provides a cooperative flight control method for a multi-rotor unmanned aerial vehicle based on a plunger pump motor, comprising: S1. Acquire the relative pose measurement data and inertial measurement data of the formation drones, and calculate the lift reference command of each drone in the current control cycle according to the formation mission requirements and safety constraints. To ensure the formation remains vertically stable and avoids altitude fluctuations, this embodiment uses relative pose measurement data and inertial measurement data in a closed loop within the same control cycle. Through step-by-step process control of altitude deviation, velocity difference, and lift command, the vertical altitude deviation is first converted into the target vertical velocity, then the velocity difference is converted into the target vertical acceleration, and finally, combining the lift execution effect of the previous control cycle with the lift margin of each UAV, the target vertical acceleration is implemented as the lift reference command for this control cycle. Process control within the executable boundary is completed through trimming and margin redistribution. The specific implementation process includes the following steps: S1-1. At the beginning of each control cycle, read the relative pose measurement data and extract the vertical component to obtain the current relative height. Then, determine the target relative height according to the formation mission requirements. Subtract the current relative height from the target relative height to obtain the height deviation. Then, read the control cycle duration and divide the height deviation by the control cycle duration to obtain the target vertical velocity. Then, limit the target vertical velocity within the allowable velocity range, thereby controlling the output of the target vertical velocity through a velocity loop process. S1-2. Read the actual vertical velocity from the inertial measurement data, and subtract the actual vertical velocity from the target vertical velocity to obtain the velocity difference; then read the control cycle duration, divide the velocity difference by the control cycle duration to obtain the target vertical acceleration; then limit the target vertical acceleration within the allowable acceleration range, thereby controlling the output of the target vertical acceleration through the acceleration loop process. S1-3: Read the vertical acceleration from the lift reference command and inertial measurement data of the previous control cycle, and subtract the vertical acceleration of the previous control cycle from the vertical acceleration of the current control cycle to obtain the change in vertical acceleration; then, use the difference between the target vertical acceleration and the vertical acceleration of the current control cycle to obtain the acceleration error, and combine it with the change in vertical acceleration to estimate the strength of the lift response, converting the acceleration error into a lift correction amount; then, superimpose the lift correction amount onto the lift reference command of the previous control cycle to obtain the initial lift reference command; then, trim the initial lift reference command according to the lift margin of each UAV, and redistribute the gap generated by the trimming according to the proportion of the remaining lift margin of the other UAVs. The process control of the lift command layer is completed through trimming and redistribution, and the lift reference command of the current control cycle is obtained and output. Through the above steps, vertical control gradually decreases from height deviation to velocity and acceleration, and then to lift command. At each level, the available margin is limited and allocated to achieve closed-loop process control from measurement to command. This ensures that the output command can be both calculated and executed, while avoiding formation oscillation caused by single-machine saturation. In practice, the three UAVs maintain the same altitude plane with the lead aircraft as a reference. In each control cycle, the altitude deviation is first calculated by relative positioning and the target vertical velocity and target vertical acceleration are obtained by process control. Then, the lift correction amount is calculated by combining the lift execution response of the previous cycle and superimposed to form the initial lift reference command. If one of the UAVs triggers a cut-off due to insufficient lift margin caused by the load, the system will distribute the gap according to the proportion of the remaining lift margin of the other two UAVs, thereby pushing the vertical error back to the allowable range and maintaining the formation stability without exceeding the limits.

[0016] S2. For each UAV in the formation, collect the plunger pump motor pressure sampling signal of the UAV and determine the starting point of the short time window from the pressure sampling signal. Generate the identification increment of the identification sequence corresponding to the UAV according to the starting point and limit the identification increment according to the current lift margin of the UAV to obtain the limited identification increment. This embodiment is used to convert the identification sequence of each UAV into an executable identification increment without disturbing the stability of the formation. The process control of alignment, limiting, gating and verification is completed in units of short time windows, so as to provide a stable input for the subsequent calculation of the downwash coupling coefficient. The pressure cycle is judged by sampling the pressure signal of the plunger pump motor to determine whether the pressure cycle is stable and to determine the starting point and cycle duration of the short time window. Then, the upper limit of the identification increment is calculated by using the lift margin and the limited identification increment is generated. Subsequently, the vertical jitter amplitude and the vertical error of the formation are used to perform gating process control on the identification increment and write the gating mark. Finally, the matching value of the previous short time window is used to perform verification process control on the intensity of the identification increment. The specific implementation process includes the following steps: S2-1. For each UAV, collect the plunger pump motor pressure sampling signal. Calculate the average pressure within the sliding window as the average value of the pressure sampling signal. Calculate the ripple amplitude as the maximum value minus the minimum value of the pressure sampling signal. Extract adjacent peak times, calculate the adjacent cycle duration sequence, and take the maximum value minus the minimum value as the cycle difference. The allowable difference is determined by the cycle difference baseline obtained from the statistics when the UAV is stably hovering and the safety constraints. Specifically, it is the baseline multiplied by the amplification factor and does not exceed the limited proportion of the control cycle duration. When the cycle difference is less than the allowable difference and the ripple amplitude is not zero, write the pressure valid flag as valid and write the cycle duration as the median of the adjacent cycle duration sequence. Otherwise, write the pressure valid flag as invalid and write the cycle duration as the cycle duration already written in the previous short time window. Output the pressure valid flag and the cycle duration. S2-2. When the pressure effective flag is valid, take the sampling point corresponding to the latest peak time as the starting point of the short time window and write it into the short time window starting point; when the pressure effective flag is invalid, use the short time window starting point written in the previous control cycle and write the rollback flag, and output the short time window starting point and rollback flag. S2-3. Read the lift margin of the UAV and calculate the upper limit of the identification increment according to the limiting rule; wherein the lift margin is obtained by subtracting the lift reference command from the upper limit of lift and is limited to not less than zero; the limiting rule takes the minimum value of the following three as the upper limit of the identification increment, the margin limit obtained by multiplying the lift margin by the safety ratio, the change limit obtained by multiplying the upper limit of the identification increment change rate by the control cycle duration, and the proportion limit obtained by multiplying the upper limit of the identification injection ratio by the amplitude of the lift reference command, and output the upper limit of the identification increment. S2-4. Read the UAV identifier and determine the unique identification sequence corresponding to the UAV identifier from the identification sequence table; map the identification sequence to the identification increment of each sub-time slice based on the starting point of the short time window, and trim the identification increment of each sub-time slice to a value whose amplitude does not exceed the upper limit of the identification increment and whose sign remains unchanged, to obtain the amplitude-limited identification increment and output it. S2-5. Within a short time window, read the vertical acceleration sequence of the inertial measurement response data and calculate the vertical jitter amplitude as the maximum absolute value of the vertical acceleration sequence minus its mean. Simultaneously, read the formation vertical error from the formation error sequence and take its absolute value to obtain the absolute value of the vertical error. Then, read the jitter allowable upper limit and the error allowable upper limit. The jitter allowable upper limit is obtained by multiplying the baseline of the vertical jitter amplitude obtained by the UAV when the identification increment is zeroed and it is hovering stably by the magnification factor. The error allowable upper limit is given by the constraint of the formation task on the vertical error. When the vertical jitter amplitude exceeds the jitter allowable upper limit or the absolute value of the vertical error exceeds the error allowable upper limit, the identification increment after amplitude limiting is set to zero and the gating flag is written as the disabled flag. Otherwise, the identification increment after amplitude limiting is retained and the gating flag is written as the enabled flag. Output the gating flag and the identification increment after amplitude limiting. S2-6. Within the previous short time window, read the identification sequence and inertial measurement response data already written to the UAV, and align the vertical acceleration sequence to the identification sequence sub-time slice according to the starting point of the short time window. Calculate the average vertical acceleration of each sub-time slice to form a response vector, and form a symbol vector from the identification sequence symbols. The matching value is the normalized correlation result of the response vector and the symbol vector minus the mean. The lowest usable matching value is obtained by statistically analyzing the historical matching values ​​of the UAV through marked short time windows, specifically taking its lower quantile and combining it with safety constraints. If the review fails, a re-examination flag is written, and the upper limit of the recognition increment is lowered according to the amplitude limiting rule. The recognition increment is then trimmed again to obtain the re-trimmed increment. Subsequently, the vertical jitter amplitude, vertical error absolute value, jitter allowable upper limit and error allowable upper limit of this short time window are used for gating judgment. When the vertical jitter amplitude exceeds the jitter allowable upper limit or the vertical error absolute value exceeds the error allowable upper limit, the re-trimmed increment is set to zero and a disable flag is written. Otherwise, an enable flag is written and the re-trimmed increment is used as the recognition increment after amplitude limiting. Through the above steps, the starting point of the short time window is determined by the pressure peak, the upper limit of the identification increment is constrained by the lift margin and the amplitude limit rule, the gating mark is written by the judgment result of the vertical jitter amplitude and the absolute value of the vertical error within or outside the allowable upper limit, and the verification is corrected by the matching value of the previous short time window. Thus, the identification injection process forms a process control link that is calculable, quantifiable, executable and traceable. In practice, when the formation enters close-range coordination, each UAV first determines the starting point and cycle duration of the short time window based on the effective pressure marker, and then generates and trims the recognition increment according to the lift margin. If the vertical jitter amplitude or the absolute value of the vertical error increases, the gating marker is written as disabled and the recognition increment is set to zero to suppress the spread of disturbance. After the disturbance returns, the gating marker is written as enabled and the recognition injection is restored. If the matching value of the previous short time window decreases, a re-examination is triggered and the upper limit of the recognition increment is lowered before trimming and gating are performed again, so as to stably output the amplitude-limited recognition increment that can be used for subsequent coupling estimation under different disturbance intensities.

[0017] S3. The lift reference command of each UAV is superimposed with the corresponding limited recognition increment to obtain the recognition lift command, and the recognition lift command is input into the plunger pump motor controller to drive the rotor to generate thrust, thereby obtaining the inertial measurement response data within a short time window. This embodiment combines the lift reference command and the limited identification increment into an executable identified lift command, and performs speed limiting under the dynamic response constraints of the plunger pump motor to ensure continuous and controllable thrust changes within a short time window. Simultaneously, inertial measurement response data is collected and aligned within the short time window to form an in-window response sequence, from which the mean vertical acceleration and vertical acceleration fluctuation are extracted as direct bases for subsequent matching statistics and gating verification. The identified lift command is first clipped at the control cycle boundary, then the speed-limited identified lift command is input into the plunger pump motor controller while simultaneously collecting the in-window response sequence. Finally, the mean and fluctuation of the in-window response sequence are calculated and written into the inertial measurement response data. The implementation process includes the following steps: S3-1. For each UAV, read the lift reference command and the limited identification increment of the current control cycle and add them together to obtain the identification lift command. At the same time, read the identification lift command saved in the previous control cycle as the command of the previous cycle. Then, calculate the difference in identification lift commands between adjacent control cycles and divide the difference by the control cycle duration to obtain the rate of change. The upper limit of the rate of change is determined by the maximum input change capability that the plunger pump motor controller can track. Specifically, the value is obtained by applying a step or ramp input to the plunger pump motor under safe ground or air conditions, recording the maximum stable tracking slope without significant overshoot and jitter, and writing it into the control parameters accordingly. When the rate of change exceeds the upper limit, the identification lift command is trimmed along the command direction of the previous cycle according to the maximum allowable increment corresponding to the upper limit of the rate of change. That is, the upper limit of the rate of change is multiplied by the control cycle duration to obtain the maximum allowable command increment, and the previous cycle command is added to the maximum allowable command increment to obtain the identification lift command after speed limitation. When the rate of change does not exceed the upper limit, the identification lift command is directly used as the identification lift command after speed limitation. S3-2. The speed-limited identification lift command is input into the plunger pump motor controller and executed within a short time window. During execution, the controller converts the identification lift command into the driving quantity of the plunger pump motor according to the internal closed loop and outputs it to the actuator. At the same time, inertial measurement response data is collected at a fixed sampling frequency within the short time window, including at least the vertical acceleration sequence and the timestamp sequence. Then, the starting point of the short time window is read, and the inertial measurement response data within the short time window is aligned according to the starting point of the short time window. Specifically, the timestamps are translated with the starting point as zero time, and the timestamps are arranged in the order of the aligned timestamps to obtain the response sequence within the window, thus forming an in-window response sequence that can be directly used for subsequent matching and statistical calculations. S3-3. Calculate the mean vertical acceleration and vertical acceleration fluctuation in the vertical acceleration sequence within the window response sequence, and write them into the inertial measurement response data: The mean vertical acceleration is the arithmetic mean of the vertical acceleration within the short time window, used to characterize the overall lift response level within the short time window; the vertical acceleration fluctuation is the maximum absolute value of the vertical acceleration sequence minus the mean vertical acceleration, used to characterize the jitter intensity within the short time window; write the mean vertical acceleration and vertical acceleration fluctuation, along with the corresponding short time window start point, into the inertial measurement response data cache for direct use in subsequent steps; Through the above steps, the lift command is first matched with the rate of change constraint of the previous cycle at the control cycle boundary to avoid overshoot or jitter caused by sudden input changes in the plunger pump motor. The inertial measurement response data within the short time window is synchronously collected and aligned according to the starting point of the short time window to form a verifiable in-window response sequence. The mean vertical acceleration and the vertical acceleration fluctuation are explicitly calculated and written into the cache, providing directly referenced fields for subsequent matching statistics and gating verification. In specific implementation, when the formation enters close-range coordination and requires the injection of identification increments, the system first adds the lift reference command and the limited identification increment to obtain the identified lift command, and then performs speed limiting and pruning of the command according to the upper limit of the rate of change that the plunger pump motor can stably track, so that the thrust change is a continuous slope. Subsequently, the command is executed within the short time window and the vertical acceleration sequence is collected, and aligned according to the starting point of the short time window to obtain the in-window response sequence. Finally, the mean vertical acceleration and the vertical acceleration fluctuation are calculated and written into the response cache, thus providing a verifiable basis for subsequent downwash coupling coefficient calculation and injection gating.

[0018] S4. For each UAV, the inertial measurement response data is aligned within the window starting point of a short time window to obtain the alignment response sequence. The alignment response sequence is then matched and statistically calculated with the identification sequences of other UAVs in the formation to obtain the set of downwash coupling coefficients. This embodiment is used to separate the influence intensity caused by the downwash action of other UAVs in the formation from the inertial measurement response data within a short time window without adding additional sensors, and to form a set of downwash coupling coefficients that can be directly used for subsequent collaborative compensation. By truncating and aligning the inertial measurement response data into an aligned response sequence at the starting point of the short time window, all subsequent calculations are performed on the same time base. Then, the aligned response sequence is used to calculate the matching value with the identification sequence of other UAVs, and the downwash coupling coefficient set is obtained by normalizing the amplitude of the identification sequence. Finally, the aligned response sequence is reconstructed using the downwash coupling coefficient set and the residual is calculated. Whether the residual falls within the allowable range is used as a self-consistency check. If necessary, the coupling coefficients are scaled down and written with a verification mark. The implementation process of this step includes the following steps: S4-1. For each UAV, read the starting point of the short time window and extract continuous data segments within the window from the sampling time corresponding to the starting point in the inertial measurement response data buffer. The length of the data segment within the window is determined by the number of periods within the window and the duration of the periods. Specifically, the duration of the window is obtained by multiplying the number of periods within the window by the duration of the periods, and then the number of sampling points within the window is calculated according to the sampling frequency. Subsequently, the extracted data segments within the window are sorted from early to late according to the sampling time, and the vertical acceleration sequence and the corresponding timestamp sequence are written together as an aligned response sequence. The timestamp is shifted with the sampling time of the starting point of the short time window as zero time, so as to obtain an aligned response sequence that corresponds one-to-one with the identification sequence. S4-2. For each UAV, calculate the matching value between the alignment response sequence of the UAV and the corresponding identification sequence of the other UAVs in the formation: First, divide the alignment response sequence into sub-time slices according to the identification sequence, calculate the average vertical acceleration of each sub-time slice to form a response vector, and then extract the code elements of the identification sequence in the same sub-time slice to form a code element vector; the matching value is the normalized correlation result of the response vector and the code element vector. The normalization method is to subtract the mean of the two vectors respectively, take the inner product and divide by the square root of the sum of their squares; then, the matching value is normalized according to the amplitude of the identification sequence to obtain the downwash coupling coefficient, where the amplitude of the identification sequence is the average absolute value or effective amplitude of the code element vector, and the downwash coupling coefficients calculated for each other UAV are collected into a downwash coupling coefficient set according to their UAV identifier; S4-3. Calculate the reconstructed response and residuals based on the downwash coupling coefficient set and the aligned response sequence: First, take the identification sequence of each other UAV and sum it by weighted downwash coupling coefficients to obtain the reconstructed response vector. Then, expand the reconstructed response vector back to the sampling points within the window according to the sub-time slice to obtain the reconstructed response sequence. Then, subtract the reconstructed response sequence from the aligned response sequence to obtain the residual sequence, and take the maximum absolute value of the residual sequence as the residual. The allowable range of the residual is obtained by statistical analysis of a short time window with the identification increment set to zero and gating enabled. Specifically, take the upper quantile of the residual and multiply it by the amplification factor. When the residual exceeds the allowable range, shrink the downwash coupling coefficient set proportionally. The shrinkage ratio is taken as the allowable range divided by the residual and limited to no more than one. At the same time, write a verification mark. When the residual does not exceed the allowable range, write a pass mark and output the downwash coupling coefficient set. The residual verification and proportional shrinkage constitute the process control of the downwash coupling coefficient set. Through the above steps, the aligned response sequence is formed on a unified short time window time base. The matching value is converted into a set of downwashed coupling coefficients according to the amplitude of the identification sequence, so that the coupling coefficients are comparable and directly usable. The residual test provides self-consistent constraints to avoid the coupling coefficients being amplified by noise, sudden disturbances or identification injection anomalies. If necessary, the estimation results are pulled back to the usable range by proportional shrinkage, thereby providing stable input for subsequent collaborative compensation. In practice, after the formation enters close formation and hovers, each UAV captures and aligns its vertical acceleration sequence at the starting point of a short time window to obtain an aligned response sequence. Then, it performs normalized correlation calculations with the individual identification sequences of other UAVs to form a set of downwash coupling coefficients. Subsequently, the response is reconstructed using this set of coupling coefficients and the residuals are calculated. If the residuals are too large, the coupling coefficients are automatically reduced and a verification mark is written. If the residuals are within the allowable range, a pass mark is written and the set of downwash coupling coefficients is output for direct use in the next step of collaborative compensation calculation.

[0019] S5. Calculate the collaborative compensation increment based on the set of downwash coupling coefficients and the lift reference command, and then superimpose the collaborative compensation increment onto the lift reference command to obtain the collaborative lift command. This embodiment is used to convert the set of downwash coupling coefficients into directly executable collaborative compensation increments, and to offset the vertical disturbances caused by downwashing of other UAVs without changing the overall lift level of the formation. The downwash coupling coefficients are used to convert the lift reference commands of other UAVs into the downwash impact on this UAV, and then the negative of the impact is used as the original compensation value. Finally, the compensation increment is clipped according to the upper limit of the compensation range and superimposed on the lift reference command to form a collaborative lift command. The upper limit of the compensation range is constrained by the lift margin and the change boundary that the controller can track, so as to ensure that the compensation increment can be effective without triggering over-limits or introducing new oscillations. The implementation process includes the following steps: S5-1. For each UAV, read the set of downwash coupling coefficients for this control cycle and the lift reference commands of other UAVs in the formation, and match the corresponding relationships one by one according to the UAV identifier; then, for each other UAV, multiply the lift reference command of the other UAV by its corresponding downwash coupling coefficient to obtain the downwash contribution value of a single UAV, and sum all the downwash contribution values ​​of a single UAV to obtain the downwash impact; the sign of the downwash impact is consistent with the downwash coupling coefficient, and it is used to characterize the combined influence intensity of the current lift of other UAVs on the vertical response of this UAV, so as to obtain the downwash impact that can be directly used for compensation calculation; S5-2. For each UAV, the downwash effect is negative to obtain the original compensation value, making the compensation direction opposite to the downwash effect direction. Then, the upper limit of the compensation amplitude is determined and the original compensation value is trimmed to obtain the collaborative compensation increment. The upper limit of the compensation amplitude is obtained by taking the smaller value of two types of constraints. One is the lift margin constraint, that is, the lift margin of the UAV is read and multiplied by the safety ratio to obtain the allowable compensation amplitude of the margin. The other is the change boundary constraint, that is, the control cycle duration and the upper limit of the compensation change rate are read and multiplied to obtain the allowable compensation amplitude of a single cycle. When the amplitude of the original compensation value exceeds the upper limit of the compensation amplitude, the original compensation value is trimmed to a value with an amplitude equal to the upper limit of the compensation amplitude and the sign remains unchanged. When the amplitude of the original compensation value does not exceed the upper limit of the compensation amplitude, the original compensation value is directly taken as the collaborative compensation increment, and thus the collaborative compensation increment is output. S5-3. For each UAV, read its lift reference command and add the cooperative compensation increment to the lift reference command to obtain the cooperative lift command. To avoid overall lift offset caused by compensation, the cooperative compensation increments of each UAV in the formation can be summed and the summation result can be evenly distributed according to the number of UAVs and then deducted from each cooperative compensation increment. Then, the summation is performed to obtain the cooperative lift command. Finally, the cooperative lift command is written into the control output buffer of this control cycle and output for the input plunger pump motor controller to be executed in the next control cycle. Through the above steps, the set of downwash coupling coefficients is clearly transformed into downwash influence, and then the collaborative compensation increment is obtained by negation and trimming, ultimately forming an executable collaborative lift command. Among them, trimming ensures that the compensation does not exceed the lift margin and change boundary, and average deduction ensures that the compensation does not introduce overall lift drift of the formation, thus stably offsetting downwash disturbances within the executable boundary. In specific implementation: when three UAVs hover in close formation, each UAV first multiplies the downwash coupling coefficient by the lift reference command of the other two UAVs and sums them to obtain the downwash influence, then negates it to form the original compensation value, and trims it according to the lift margin and single-cycle change boundary to obtain the collaborative compensation increment. Finally, the collaborative compensation increment is superimposed on the lift reference command to generate a collaborative lift command and output it, thereby offsetting the vertical deviation caused by downwash in a continuous and controllable manner in the next control cycle and maintaining the formation height stability.

[0020] S6. Input the coordinated lift command of each UAV into the corresponding plunger pump motor controller and execute it in the next control cycle. At the same time, update the number of cycles in the window and identify the incremental limit rule according to the set of downwash coupling coefficients and the formation error sequence to obtain the updated parameters. This embodiment is used to continuously update the time window settings and identification increment limiting rules upon which subsequent control depends when a coordinated lift command has been formed and is about to be executed. This allows the system to automatically adjust the injection intensity and statistical window according to changes in formation coupling strength and formation error, avoiding the introduction of new oscillations by using old parameters when coupling changes abruptly or errors amplify. By executing the coordinated lift command in the next control cycle, the set of downwash coupling coefficients and formation error sequence of the current control cycle are simultaneously solidified into the current cycle record. The coupling change is then obtained using the current cycle record and the previous control cycle record, and the error fluctuation is obtained from the formation error sequence of the current cycle. Both are used to determine the number of cycles within the window and to calculate the duration within the window. Finally, the updated number of cycles within the window, coupling change, and lift margin are used to update the identification increment limiting rules and output updated parameters to perform process control on the identification injection intensity and statistical window, ensuring that the boundaries of subsequent identification increment generation and gating verification are executable and match the current operating conditions. The implementation process includes the following steps: S6-1. At the control cycle switching moment, the cooperative lift command calculated in this control cycle is written into the control output buffer, and at the beginning of the next control cycle, the cooperative lift command is input into the corresponding plunger pump motor controller for execution. The plunger pump motor controller outputs the drive amount to change the rotor thrust accordingly. At the same time, before the end of this control cycle, the set of downwash coupling coefficients and the formation error sequence of this control cycle are read and written into the recording area along with the control cycle number and timestamp to form the record of this cycle. The formation error sequence shall at least include the sampling sequence of the vertical height deviation or speed deviation within this control cycle to ensure that the error fluctuation can be calculated subsequently. S6-2. Calculate the change in the set of downwash coupling coefficients based on the records of the current cycle and the previous control cycle, and calculate the error fluctuation based on the formation error sequence of the current cycle: The change is calculated by subtracting the downwash coupling coefficients of the same pair of UAVs and taking the absolute value, then taking the maximum or average value of all differences within the set as the change; the error fluctuation is calculated by first calculating the mean of the formation error sequence of the current cycle, then using the maximum absolute value of the error sequence minus the mean as the error fluctuation; subsequently, update the number of cycles within the window based on the change and the error fluctuation. The new rules are implemented using a tiered increase / decrease method. When the change or error fluctuation increases, the number of periods within the window is reduced to shorten the statistical window and improve response speed. When both the change and error fluctuation are small, the number of periods within the window is increased to extend the statistical window and improve estimation stability. The number of periods within the window is always limited to between the minimum and maximum number of periods within the window. Finally, the period duration is read and multiplied by the number of periods within the window to calculate the window duration, thus obtaining the number of periods within the window and the window duration. The process control of the statistical window is achieved by tiered increases and decreases in the number of periods within the window. S6-3. Update the identification increment limiting rule and form update parameters based on the number of cycles within the window, the amount of change, and the lift margin: First, read the lift margin of the UAV and multiply it by a safety ratio to obtain the margin limiting benchmark; then, reduce the margin limiting benchmark according to the amount of change, with the reduction rule being that the larger the amount of change, the larger the reduction ratio, in order to reduce the identification injection intensity when the coupling changes rapidly; then, perform window correction on the reduced limiting benchmark according to the number of cycles within the window, with the window correction rule being that the smaller the number of cycles within the window, the smaller the allowable injection amplitude per cycle, in order to avoid the identification increment being too strong under short window conditions, which would lead to matching statistical distortion; write the obtained limiting result into the identification increment limiting rule, and write the number of cycles within the window, the duration within the window, and the updated identification increment limiting rule together into the update parameter area and output the update parameters for direct use in the determination of the short time window and identification increment limiting in the next control cycle; Through the above steps, the coordinated lift command is executed according to the predetermined link in the next control cycle. At the same time, the set of downwash coupling coefficients and the formation error sequence are solidified into a traceable record for this cycle. The coupling change and error fluctuation are explicitly calculated and used to update the number of cycles and the duration within the window, so that the statistical window can be shortened to improve response when unstable and extended to improve estimation reliability when stable. The identification increment limit rule is updated synchronously with the lift margin, coupling change and window size, so that subsequent identification increment injections are always within the executable boundary and match the current operating conditions. In practical implementation, when the formation enters close-range coordination, each control cycle first executes the coordinated lift command generated in the previous cycle, and at the same time records the set of downwash coupling coefficients and the formation vertical error sequence of the current cycle. If the change in downwash coupling coefficients suddenly increases or the formation error fluctuation increases, the system immediately reduces the number of cycles in the window and shortens the window duration, while tightening the identification increment limit rule to reduce the injection intensity. After the coupling change and error fluctuation subside, the number of cycles in the window is gradually increased and the limit is relaxed, so as to maintain the continuous stability of identification injection, coupling estimation and coordinated compensation under different operating conditions.

[0021] Working Principle: This solution addresses the issue of mutual downwash interference in close-range multi-rotor formations. First, relative attitude and inertial measurement are used to calculate the lift reference command for each UAV, ensuring the formation meets mission and safety constraints in the vertical direction. Then, each UAV determines the start point and cycle length of a short time window from the plunger pump motor pressure sampling signal. The identification sequence bound to the UAV identifier is converted into an identification increment, which is then trimmed into an executable, limited identification increment based on lift margin and limiting rules. Simultaneously, vertical jitter and formation error are used for gating, and the matching value from the previous short time window is used for verification, ensuring that the identification injection is both stable and traceable. Finally, the lift reference command is... The recognition lift command is generated by superimposing the amplitude limit recognition increment, and then input into the plunger pump motor controller for execution after speed limiting according to the upper limit of the rate of change. Inertial measurement response data is synchronously collected and aligned within a short time window. The aligned response sequence is then matched and statistically analyzed with other UAV recognition sequences to obtain the downwash coupling coefficient set, and the convergence coefficient is verified by residual self-consistency. Finally, the cooperative compensation increment is calculated based on the downwash coupling coefficient set, superimposed on the lift reference command to generate a cooperative lift command, and executed in the next control cycle. At the same time, the number of cycles in the window and the recognition increment amplitude limit rule are updated according to the coupling change and formation error fluctuation, so that the subsequent recognition injection and coupling estimation always fit the current working conditions. For example, when three drones perform close-range hovering formation in a narrow airspace, relying solely on individual altitude control often results in repeated fluctuations, with one drone rising and another sinking. In this solution, the system first generates individual lift reference commands for each of the three drones to maintain formation. Simultaneously, each drone uses its own pressure waveform to determine the same type of time reference, injecting a recognition increment within a short time window, the amplitude of which is limited by lift margin. When jitter or formation error increases, the system immediately gating to zero to prevent further confusion. After the short time window ends, each drone compares its aligned vertical acceleration response with that of the other two drones. The system matches the identification sequence to obtain the downwash coupling strength of the other two aircraft to the main aircraft, and uses residual checks to shrink inconsistent estimates back to the usable range. Then, it calculates the compensation increment based on the coupling strength, appropriately increases the lift of the aircraft more affected by the downwash, and appropriately decreases the lift of the aircraft less affected, so as to cancel out the downwash disturbance in the next control cycle. If a sudden approach causes the coupling change to be faster or the error fluctuation to increase, the system will automatically shorten the number of cycles in the window and tighten the amplitude limit rules. After the operating conditions return to stability, the rules will be gradually relaxed so that the formation can fly stably and cooperatively under different distances and disturbances.

[0022] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A cooperative flight control method for a multi-rotor unmanned aerial vehicle based on a plunger pump motor, characterized in that, include: S1. Acquire the relative pose measurement data and inertial measurement data of the formation drones, and calculate the lift reference command of each drone in the current control cycle according to the formation mission requirements and safety constraints. S2. For each UAV in the formation, collect the plunger pump motor pressure sampling signal of the UAV and determine the starting point of the short time window from the pressure sampling signal. Generate the identification increment of the identification sequence corresponding to the UAV according to the starting point and limit the identification increment according to the current lift margin of the UAV to obtain the limited identification increment. For each UAV, the pressure sampling signal of the plunger pump motor is collected. The average pressure, ripple amplitude and period difference are calculated in the sliding window. When the period difference is less than the allowable difference and the ripple amplitude is not zero, the pressure valid flag and period duration are written and the pressure valid flag and period duration are output. When the pressure valid flag is valid, the starting point of the short time window is determined based on the adjacent peak times of the pressure sampling signal and written into the short time window starting point. When the pressure valid flag is invalid, the starting point of the short time window of the previous control cycle is used and the rollback flag is written into the short time window. The starting point of the short time window and the rollback flag are output. Read the lift margin of the drone, convert the lift margin into the upper limit of recognition increment according to the limit rule, and output the upper limit of recognition increment; Read the drone's identifier and determine the corresponding identification sequence. Based on the starting point of the short time window, map the identification sequence into an identification increment. Then, trim the identification increment according to the upper limit of the identification increment to obtain the narrowed identification increment. Output the narrowed identification increment. The vertical jitter amplitude is calculated based on the formation error sequence and inertial measurement response data. The vertical jitter amplitude is then gating according to the limiting rule. If the gating fails, the recognition increment after limiting is set to zero and a disabled flag is written. If the gating passes, an enabled flag is written and the recognition increment after limiting is retained. The gating flag and the recognition increment after limiting are then output. The matching value is calculated based on the recognition sequence and inertial measurement response data of the previous short time window. The matching value is then checked and judged according to the amplitude limiting rule. If the check fails, a re-check mark is written and the upper limit of the recognition increment is updated according to the amplitude limiting rule. The recognition increment is then trimmed again and gating is performed to update the amplitude-limited recognition increment. If the check passes, a pass mark is written and the amplitude-limited recognition increment is output. S3. The lift reference command of each UAV is superimposed with the corresponding limited recognition increment to obtain the recognition lift command, and the recognition lift command is input into the plunger pump motor controller to drive the rotor to generate thrust, thereby obtaining the inertial measurement response data within a short time window. S4. For each UAV, the inertial measurement response data is aligned within the window starting point of a short time window to obtain the alignment response sequence. The alignment response sequence is then matched and statistically calculated with the identification sequences of other UAVs in the formation to obtain the set of downwash coupling coefficients. S5. Calculate the collaborative compensation increment based on the set of downwash coupling coefficients and the lift reference command, and then superimpose the collaborative compensation increment onto the lift reference command to obtain the collaborative lift command. S6. Input the coordinated lift command of each UAV into the corresponding plunger pump motor controller and execute it in the next control cycle. At the same time, update the number of cycles in the window and identify the incremental limit rule according to the set of downwash coupling coefficients and the formation error sequence to obtain the updated parameters.

2. The method for cooperative flight control of a multi-rotor UAV based on a plunger pump motor according to claim 1, characterized in that: S1 includes: S1-1. Calculate the vertical height deviation of the formation based on the relative pose measurement data, divide the height deviation by the current control cycle duration to obtain the target vertical velocity, and limit the target vertical velocity within the speed range allowed by the safety constraints, and output the target vertical velocity. S1-2. Calculate the velocity difference based on the target vertical velocity and the actual vertical velocity in the inertial measurement data, divide the velocity difference by the current control cycle duration to obtain the target vertical acceleration, and limit the target vertical acceleration within the acceleration range allowed by the safety constraints, and output the target vertical acceleration. S1-3. Calculate the lift correction amount based on the lift reference command of the previous control cycle, the change in vertical acceleration in the inertial measurement data, and the lift margin of each UAV. Then, superimpose the lift correction amount onto the lift reference command of the previous control cycle and perform trimming and margin redistribution according to the lift margin to obtain the lift reference command of the current control cycle, and output the lift reference command.

3. The method for cooperative flight control of a multi-rotor UAV based on a plunger pump motor according to claim 2, characterized in that: S3 includes: S3-1. For each UAV, add the lift reference command to the recognition increment after limiting to obtain the recognition lift command. Divide the difference of the recognition lift command between adjacent control cycles by the control cycle duration to obtain the rate of change. Then, trim the rate of change according to the upper limit of the rate of change to obtain the speed-limited recognition lift command. S3-2. Input the speed-limited identification lift command into the plunger pump motor controller and execute it within a short time window. At the same time, collect inertial measurement response data within the short time window and align it with the starting point of the short time window to form a response sequence within the window. S3-3. Calculate the mean vertical acceleration and the vertical acceleration fluctuation for the response sequence within the window, and write the mean vertical acceleration and the vertical acceleration fluctuation into the inertial measurement response number.

4. The method for cooperative flight control of a multi-rotor UAV based on a plunger pump motor according to claim 3, characterized in that: S4 includes: S4-1. For each UAV, inertial measurement response data is extracted at the starting point of a short time window and arranged in the order of sampling time to obtain an aligned response sequence; S4-2. Calculate the matching value between the alignment response sequence and the corresponding identification sequence of each of the other UAVs in the formation, and normalize the matching value according to the amplitude of the identification sequence to obtain the set of downwash coupling coefficients. S4-3. Calculate the reconstructed response based on the set of downwashed coupling coefficients and the aligned response sequence, and obtain the residual by subtracting from the aligned response sequence. When the residual exceeds the allowable range, shrink the set of downwashed coupling coefficients proportionally and write a verification mark. When the residual does not exceed the allowable range, write a pass mark and output the set of downwashed coupling coefficients.

5. The method for cooperative flight control of a multi-rotor UAV based on a plunger pump motor according to claim 4, characterized in that: S5 includes: S5-1. For each UAV, multiply the lift reference command of other UAVs in the formation by the corresponding downwash coupling coefficient and sum them to obtain the downwash influence amount. S5-2. For each drone, the negative value of the downwash effect is used to obtain the original compensation value, and the original compensation value is trimmed according to the upper limit of the compensation range to obtain the collaborative compensation increment. S5-3. For each UAV, the collaborative compensation increment is superimposed on the lift reference command to obtain the collaborative lift command and output it.

6. The method for cooperative flight control of a multi-rotor UAV based on a plunger pump motor according to claim 5, characterized in that: S6 includes: S6-1. Input the coordinated lift command into the plunger pump motor controller and execute it in the next control cycle. At the same time, record the set of downwash coupling coefficients and the formation error sequence of this control cycle to obtain the record of this cycle. S6-2. Calculate the change in the set of washing coupling coefficients based on the current cycle record and the previous control cycle record, and calculate the error fluctuation based on the current cycle formation error sequence. Update the number of cycles within the window based on the change and fluctuation, and calculate the duration within the window based on the cycle duration and the number of cycles within the window. S6-3. Based on the number of cycles within the window, the amount of change, and the lift margin, update and identify the incremental limiting rules and form update parameters, then output the update parameters.

Images