Intelligent control method of unmanned aerial vehicle holder stability augmentation micro motor

Abstract

The invention relates to the technical field of pan-tilt motor control, in particular to an intelligent control method for an unmanned aerial vehicle pan-tilt stability augmentation micro motor. Synchronously acquiring operation parameter time sequence data, image data, flight attitude data, environment wind speed, pan-tilt motor power and rotating speed of the unmanned aerial vehicle; constructing a state fluctuation index (based on operation parameter time sequence data) and a flight attitude deviation degree (fusing wind speed change and flight attitude deviation), and further generating a wind resistance capability index to evaluate the influence degree of the environment on the unmanned aerial vehicle; meanwhile, on the basis of the change characteristics of the parting attitude data, the compensation effect of the pan-tilt motor is quantified by combining the image fuzziness, the power of the motor and the rotating speed fluctuation; finally, PID parameters are dynamically optimized based on double indexes of the wind resistance and the compensation effect, scene adaptation of a pan-tilt motor control strategy is achieved, the stability of the unmanned aerial vehicle is improved in a complex wind field, and the image data quality is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of pan / tilt motor control, and in particular to an intelligent control method for a pan / tilt stabilization micro motor of an unmanned aerial vehicle (UAV) apparatus. Background Art

[0002] In recent years, with the rapid development of drone technology, drones have become increasingly widely used in fields such as agricultural monitoring, environmental monitoring, logistics distribution, and film and television production. A drone's stability and control accuracy directly impact its flight safety and mission performance. Therefore, the stability of the gimbal system is a key element in ensuring mission quality. Micromotors are one of the core components of a gimbal system. Their primary task is to control motion through power output, compensating for gimbal attitude fluctuations.

[0003] Existing technologies usually use intelligent control methods, such as PID control, to adaptively adjust the output power of micro motors based on real-time data. However, they often use fixed preset PID parameters. In complex situations such as rapid movement of drones or external interference, the preset PID parameters will result in a long response time and an inability to quickly correct errors. This will not only lead to unstable drone status, but also waste motor energy and affect flight endurance. Summary of the Invention

[0004] In order to solve the technical problem that in the complex situation of rapid movement of drones or external interference, the preset PID parameters will lead to a long response time and the inability to quickly correct errors, which will not only lead to unstable drone state but also cause energy waste of the motor and affect the flight life, the purpose of the present invention is to provide an intelligent control method for drone gimbal stabilization micro motor, and the technical solutions adopted are as follows:

[0005] Obtaining the time series data of the UAV's operating parameters, actual flight attitude data at each moment, and captured image data. The operating parameters include angular velocity and acceleration in each direction; obtaining the time series data of the wind speed in the UAV's environment; obtaining the time series data of the output power and speed of the gimbal motor;

[0006] The drone's state fluctuation index is determined based on the changing trend and fluctuation characteristics of angular velocity and the fluctuation differences between accelerations in all directions. The drone's flight attitude deviation is determined based on the changing trend and numerical differences of wind speed values, as well as the deviation between the actual flight attitude data at each moment and the preset flight attitude data. The drone's wind resistance index is determined based on the flight attitude deviation and state fluctuation index.

[0007] Based on the variation characteristics of the actual flight attitude data of the UAV, the fuzzy variation between the image data is analyzed. Combined with the fluctuation characteristics of the output power and speed of the gimbal motor, the compensation effect index of the gimbal motor is determined.

[0008] The preset PID parameters are adjusted based on the compensation effect index of the gimbal motor and the wind resistance index of the drone to control the output power of the gimbal motor.

[0009] Furthermore, the method for obtaining the state fluctuation index includes:

[0010] In the angular velocity time series data, analyze the changing trend and fluctuation characteristics of the angular velocity to determine the rapid swing index of the drone;

[0011] In the acceleration time series data in each direction, the variance of all accelerations is used as the change fluctuation factor corresponding to each direction;

[0012] Combine all directions in pairs to obtain all direction combinations. In each direction combination, calculate the absolute value of the difference between the change fluctuation factors corresponding to the two directions as the state change factor. Normalize the mean of the state change factors of all direction combinations to obtain the value as the state fluctuation parameter of the drone.

[0013] The product of the rapid swing index and the state fluctuation parameter is normalized to a value which is used as the state fluctuation index of the UAV.

[0014] Furthermore, the method for obtaining the fast swing index includes:

[0015] Obtaining a data curve corresponding to the angular velocity time series data as an angular velocity data curve, and obtaining a slope value at each angular velocity on the angular velocity data curve;

[0016] Calculate the absolute value of the difference between the slope values ​​of each two adjacent angular velocities as the trend change factor, and take the average of all trend change factors as the trend change parameter;

[0017] Obtain all extreme points on the angular velocity data curve;

[0018] The time interval between each two adjacent extreme points is taken as the interval factor, and the mean of all interval factors is taken as the time interval parameter;

[0019] The value after normalizing the ratio of the trend change parameter to the time interval parameter is used as the rapid swing index of the UAV.

[0020] Furthermore, the method for obtaining the flight attitude deviation includes:

[0021] In the wind speed time series data, the time series is divided based on the change trend and numerical difference of the wind speed value to obtain the time period of rapid change and the time period of slow change;

[0022] In each of the drastic change time periods, between the actual flight attitude data and the preset flight attitude data at each moment, the absolute value of the difference between the same type of data values ​​is used as the attitude deviation factor, and the sum of the attitude deviation factors corresponding to all types of data values ​​is used as the attitude deviation parameter;

[0023] The mean of the attitude deviation parameters at all moments in all drastic change periods is taken as the attitude deviation factor;

[0024] Calculating the absolute value of the difference between the mean of all wind speed values ​​in all the drastic change time periods and the mean of all wind speed values ​​in all the slow change time periods as the wind speed change factor;

[0025] The product of the attitude deviation factor and the wind speed change factor is normalized to a value which is used as the flight attitude deviation degree of the UAV.

[0026] Furthermore, in the wind speed time series data, the time series is divided based on the change trend and numerical difference of the wind speed value to obtain a time period of rapid change and a time period of slow change, including:

[0027] Obtaining a data curve of the wind speed time series data as a wind speed data curve, and obtaining an absolute value of a slope at each wind speed value on the wind speed data curve;

[0028] Sort all the absolute values ​​of the slopes in order to obtain an ascending sequence;

[0029] In the ascending sequence, a difference sequence of all slope absolute values ​​is calculated, and in the difference sequence, two slope absolute values ​​corresponding to the maximum value in the ascending sequence are used as mark values;

[0030] In the ascending sequence, the wind speed value corresponding to the maximum value of the two marked values ​​and the absolute value of the slope thereafter on the wind speed data curve is taken as the first wind speed value, and the wind speed value corresponding to the minimum value of the two marked values ​​and the absolute value of the slope thereafter on the wind speed data curve is taken as the second wind speed value;

[0031] On the wind speed data curve, a time period corresponding to a subsequence of first wind speed values ​​that are continuous in time series is used as a time period of rapid change, and a time period corresponding to a subsequence of second wind speed values ​​that are continuous in time series is used as a time period of slow change.

[0032] Get all the time periods of rapid change and slow change.

[0033] Furthermore, the method for obtaining the wind resistance index includes:

[0034] The product of the state fluctuation index and the flight attitude deviation is negatively correlated and normalized to a value obtained as the wind resistance index of the UAV.

[0035] Furthermore, the method for obtaining the compensation effect indicator includes:

[0036] Between any two adjacent moments, compare the data values ​​of the actual flight attitude data of the UAV under the same category. If there is a group of data values ​​that are different, then the two adjacent moments are regarded as a group of change moments.

[0037] Fusion analysis of the blur changes between the two image data in all time change groups to determine the first loss factor;

[0038] Multiplying the variance of the output power values ​​in the output power time series data by the variance of the speed values ​​in the speed time series data, and normalizing the obtained product as a second loss factor;

[0039] The value obtained by performing negative correlation mapping on the product of the first loss factor and the second loss factor is used as the compensation effect indicator of the pan / tilt motor.

[0040] Furthermore, the method for obtaining the first loss factor includes:

[0041] In each group of changing moments, the image data at each moment is processed based on the Canny operator to obtain the gradient amplitude of the pixel points. The variance of the gradient amplitudes of all pixel points is used as the blur amount. The absolute value of the difference between the blur amounts of the image data at two moments is used as the blur difference factor. The product of the value after negative correlation mapping of the blur amount of the image data at the previous moment in the temporal sequence and the blur difference factor is used as the blur eigenvalue.

[0042] The normalized value of the mean of the fuzzy eigenvalues ​​of all change moment groups is used as the first loss factor.

[0043] Furthermore, the method for obtaining the preset PID parameters includes:

[0044] The preset PID parameters include an initial proportional gain coefficient, an initial integral gain coefficient, and an initial differential gain coefficient;

[0045] The initial proportional gain coefficient, initial integral gain coefficient, and initial differential gain coefficient are obtained based on the Ziegler-Nichols method.

[0046] Furthermore, the preset PID parameters are adjusted based on the compensation effect index of the pan-tilt motor and the wind resistance index of the UAV to control the output power of the pan-tilt motor, including:

[0047] The compensation effect index is negatively correlated and normalized, and the result is used as a first adjustment factor;

[0048] The wind resistance index of the UAV is negatively correlated and normalized, and the value obtained is used as the second adjustment factor;

[0049] Taking the sum of the first adjustment factor and the second adjustment factor as an adjustment parameter;

[0050] Multiplying the adjustment parameter by the initial proportional gain coefficient of the PID controller to obtain an adjustment proportional gain coefficient;

[0051] The output power of the motor is controlled by PID based on the adjusted proportional gain coefficient, the initial integral gain coefficient and the initial differential gain coefficient.

[0052] The present invention has the following beneficial effects:

[0053] By synchronously collecting the drone's operating parameters (angular velocity, acceleration), flight attitude data, image data, ambient wind speed, and gimbal motor output power and speed time series data, a comprehensive understanding of the flight environment and load status is achieved, providing a data foundation for subsequent precise control. In complex environments, the preset PID parameters can result in long response times, making it difficult to adjust errors in a timely manner. Therefore, the environmental impact of the drone can be analyzed and the preset PID parameters adjusted based on the drone's operating status. First, a state fluctuation index is calculated based on the difference between angular velocity trends and acceleration fluctuations. This accurately characterizes the degree of vibration and attitude disturbance within the drone. This state fluctuation index provides a basis for dynamic compensation in gimbal motor control. Wind speed variations and attitude deviation data are combined to quantify flight attitude deviation, reflecting the impact of environmental disturbances on drone stability in real time. Determining flight attitude deviation allows subsequent gimbal motor control strategies to better adapt to wind disturbance trends. By integrating state fluctuations and attitude deviation, a wind resistance index is generated, dynamically assessing the drone's stability in complex wind fields. Furthermore, based on the changing characteristics of actual flight attitude data, the effectiveness of micromotor jitter compensation can be analyzed by combining image blur changes with motor power / speed fluctuations to quantify the compensation effect. Finally, PID parameters are optimized based on wind resistance and compensation effect indicators to achieve scenario-specific adaptation of the micromotor control strategy: the compensation effect indicator constrains the output power of the gimbal motor, reducing average power consumption while ensuring gimbal stability; the wind resistance indicator enables the drone gimbal control strategy to adapt to complex environments. This allows for rapid error correction, ensuring drone stability while reducing motor energy waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 A flowchart of an intelligent control method for a micro-motor for stabilizing a gimbal of an unmanned aerial vehicle provided by one embodiment of the present invention;

[0056] Figure 2 A flow chart of a method for obtaining a state fluctuation index provided by one embodiment of the present invention;

[0057] Figure 3 A flow chart of a method for obtaining a flight attitude deviation provided by one embodiment of the present invention;

[0058] Figure 4 A flowchart of a method for obtaining a compensation effect indicator provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effectiveness of the intelligent control method for a drone gimbal stabilization micromotor proposed by the present invention. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0060] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0061] The specific scheme of the intelligent control method of the drone gimbal stabilization micro motor provided by the present invention is described in detail below with reference to the accompanying drawings.

[0062] See also Figure 1 , which shows a method flow chart of an intelligent control method for a drone gimbal stabilization micro motor provided by one embodiment of the present invention, the method comprising the following steps:

[0063] Step S1: Obtain the time series data of the UAV's operating parameters, the actual flight attitude data at each moment, and the captured image data. The operating parameters include angular velocity and acceleration in each direction; obtain the time series data of the wind speed in the UAV's environment; obtain the time series data of the output power and speed of the gimbal motor.

[0064] With the widespread application of drone technology in aerial photography, surveying and mapping, security and other fields, the stability of the gimbal system has become a core element to ensure the quality of mission execution. Therefore, gimbal stabilization technology has come into being. The gimbal is a supporting structure installed on the drone. Its core component, the micro motor, can control motion through output power, thereby compensating for the drone's attitude fluctuations, reducing external interference to the drone, and ensuring a stable image.

[0065] The existing technology usually adjusts the gimbal motor output through a PID controller to compensate for the fluctuations in the flight attitude of the drone. However, it often uses fixed preset PID parameters. In complex environments or when the drone moves rapidly, the preset PID parameters will result in a long response time and an inability to quickly correct errors. Therefore, in this embodiment of the present invention, by synchronously collecting the drone's operating parameter time series data (operating parameters include angular velocity and acceleration in various directions), flight attitude data at various moments, and captured image data, and the wind speed time series data of the drone's environment, it is possible to fully perceive the flight environment and load status, providing a data basis for subsequent precise control.

[0066] Specifically, an inertial measurement unit (IMU) can be deployed at the center of gravity of the drone to capture angular velocity time series data and acceleration time series data in three directions (X-axis, Y-axis, and Z-axis) at a sampling frequency of 1kHz. The angular velocity time series data is used to detect the rotation changes of the drone, and the acceleration time series data is used to provide dynamic information about the drone in various directions. Then, based on the angular velocity and acceleration at each moment, the actual flight attitude data (pitch angle, roll angle, and heading angle, etc.) of the drone at each moment can be calculated. At the same time, the camera equipment on the drone is used to capture image data, and the sampling frequency of the image data can also be set to 1kHz. A three-dimensional ultrasonic anemometer is installed on the top of the drone to measure the wind speed time series data of the drone's environment, and the sampling frequency is also set to 1kHz. At this point, the drone's related operating parameter time series data, actual flight attitude data, captured image data, and data on the environment can be obtained.

[0067] The Hall sensor can be used to monitor the power timing data and speed timing data of the micro pan-tilt motor to reflect the real-time working status of the pan-tilt. Similarly, the sampling frequency can be set to 1kHz.

[0068] Select a suitable microcontroller as the core unit for data acquisition and processing. Depending on the type of the aforementioned sensors, connect them to the microcontroller through appropriate communication interfaces to ensure real-time data transmission. Before starting data acquisition, calibrate the sensors to ensure measurement accuracy. During the data acquisition process, apply filtering techniques, such as low-pass filtering or Kalman filtering, to remove unnecessary noise and improve data accuracy. Store the processed data in real time in the microcontroller memory for subsequent analysis and use.

[0069] It should be noted that the collection equipment and sampling frequency of the aforementioned various data can be adjusted according to the implementation scenario and are not limited here, but all types of data need to be collected synchronously and the sampling frequency needs to be kept consistent; in this embodiment of the present invention, the length of the time series data is set to 5 minutes, and the specific length can also be adjusted according to the implementation scenario and is not limited here.

[0070] Step S2: Determine the state fluctuation index of the UAV based on the changing trend and fluctuation characteristics of the angular velocity and the fluctuation difference between the accelerations in all directions; determine the flight attitude deviation of the UAV based on the changing trend and numerical difference of the wind speed value and the deviation between the actual flight attitude data and the preset flight attitude data at each moment; determine the wind resistance index of the UAV based on the flight attitude deviation and the state fluctuation index.

[0071] By identifying unstable states during drone flight, it's possible to better predict and respond to various disturbances (such as wind speed variations and airflow disturbances). This allows adjustments to the drone's gimbal to mitigate some of these instabilities, improving the quality of aerial video and images while effectively reducing micromotor energy consumption. Therefore, the degree of attitude fluctuation caused by mechanical vibration, turbulence, and other factors can be characterized by the angular velocity trend, fluctuation, and acceleration differences, providing internal state feedback for gimbal motor control. Furthermore, wind speed variations and flight attitude deviations can be combined to assess the intensity of external environmental disturbances affecting the drone's attitude. Finally, the degree of state fluctuation and flight attitude deviation are combined to generate a wind resistance index for the drone, reflecting its stability margin in the current environment. This provides quantitative input for subsequent adaptive adjustment of PID parameters, avoiding the risk of untimely error response caused by fixed parameters.

[0072] During the flight of a drone, changes in angular velocity indicate that the drone is experiencing an unstable flight phase. If the acceleration of the drone changes in all directions, especially when the acceleration changes suddenly due to external interference, this situation will place a high demand on stabilization micromotors. Therefore, we first analyze the change trends and fluctuation differences of the drone's angular velocity time series data and acceleration time series data to determine the drone's state fluctuation indicators.

[0073] Preferably, in one embodiment of the present invention, the method for obtaining the state fluctuation index includes:

[0074] See also Figure 2 , which shows a flow chart of a method for obtaining a state fluctuation index in one embodiment of the present invention, the method comprising the following steps:

[0075] Step S201: Analyze the changing trend and fluctuation characteristics of the angular velocity in the angular velocity time series data to determine the rapid swing index of the UAV.

[0076] Angular velocity is the instantaneous rate of attitude change and can directly describe the rotation speed of the drone. When the drone encounters a gust of wind, the angular velocity will suddenly change, and the sudden change in angular velocity will occur earlier, so the swaying state of the drone can be captured more timely. At the same time, since the slope can reflect the changing trend of the data, the data curve corresponding to the angular velocity time series data is obtained (the least squares method can be used, which is a well-known technology and the process is not described in detail) as the angular velocity data curve, and the slope value at each angular velocity is obtained on the angular velocity data curve.

[0077] Then, the absolute value of the difference in slope between each two adjacent angular velocities is calculated as the trend change factor. When the trend change factor is larger, it means that the change trend between the two adjacent angular velocities has changed significantly. Therefore, the mean of all trend change factors is calculated as the trend change parameter. Similarly, the larger the trend change parameter, the larger and more frequent the angular velocity mutation characteristics of the UAV are, and the changes are more complex, indicating that the UAV has a tendency to swing rapidly.

[0078] Extreme points often represent local maxima or minima, which are extreme critical features, that is, critical points for changes in the direction of angular velocity. On the angular velocity data curve, all extreme points are obtained, and the time interval between each two adjacent extreme points is used as the interval factor. The smaller the interval factor, the shorter the interval of changes in the direction of angular velocity. The mean of all interval factors is used as the time interval parameter. Similarly, the smaller the time interval parameter, the shorter the swing period and the more frequent the changes in angular velocity.

[0079] Based on the above analysis, it can be seen that the time interval parameter is negatively correlated with the rapid swing degree of the drone, and the trend change parameter is positively correlated with the rapid swing degree of the drone. Therefore, in this embodiment of the present invention, the ratio of the trend change parameter to the time interval parameter is normalized and used as the rapid swing index of the drone. At this time, the larger the rapid swing index is, the more frequent the angular velocity changes of the drone in the current period, and the flight state is in a relatively unstable stage, which will require a higher stabilization micromotor. Normalization is a technical means well known to those skilled in the art. The normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0080] Step S202: Analyze the fluctuation difference characteristics between the acceleration time series data in all directions to determine the state fluctuation parameters of the UAV.

[0081] The acceleration data reflects the translational vibration of the drone in three orthogonal directions. Therefore, in the acceleration time series data in each direction, the variance of all accelerations is used as the variation fluctuation factor corresponding to each direction. The larger the variation fluctuation factor, the greater the vibration in each direction.

[0082] However, the change fluctuation factor in a single direction can only reflect local vibration, while the stability of the drone needs to consider the coordinated fluctuation in three-dimensional space. Therefore, all directions are combined in pairs to obtain all non-repeating direction combinations. Then, in each direction combination, the absolute value of the difference between the change fluctuation factors corresponding to the two directions is calculated as the state change factor. The larger the state change factor, the greater the deviation between the vibrations in the two directions, the stronger the asymmetry, and the worse the stability of the drone. Therefore, the mean value of the state change factor of all direction combinations is normalized and used as the state fluctuation parameter of the drone. At this time, the larger the state fluctuation parameter, the worse the flight stability of the drone in the current period. Normalization is a technical means well known to those skilled in the art. The normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0083] Step S203: The rapid swing index and the state fluctuation parameter of the UAV are integrated to obtain the state fluctuation index of the UAV.

[0084] The analysis process in the above steps shows that both the rapid sway index and the state fluctuation parameter are negatively correlated with the drone's state stability. Therefore, the product of the rapid sway index and the state fluctuation parameter is normalized and used as the drone's state fluctuation index. The larger the state fluctuation index, the worse the drone's state stability during flight and the greater the need for a stabilization micromotor. Normalization is a technical method well known to those skilled in the art. The normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited here.

[0085] After analyzing the unstable motion state that may occur during the flight of the drone, considering that the unstable motion state may be caused by environmental factors, such as strong wind environment will cause the drone and its gimbal to shake violently during flight, the changing trend and numerical difference of the wind speed value can be analyzed, and the deviation between the actual flight attitude data and the preset flight attitude data can be compared to obtain the flight attitude deviation of the drone, thereby quantifying the interference of environmental factors during the flight of the drone.

[0086] Preferably, in one embodiment of the present invention, the method for obtaining the flight attitude deviation includes:

[0087] See also Figure 3, which shows a flow chart of a method for obtaining a flight attitude deviation in one embodiment of the present invention, the method comprising the following steps:

[0088] Step S211: In the wind speed time series data, the time series is divided based on the change trend and numerical difference of the wind speed value to obtain a time period of rapid change and a time period of slow change.

[0089] Wind speeds of varying degrees have different effects on the flight attitude of the drone. Under drastic wind speed changes, the drone is more likely to deviate from its flight attitude. Therefore, in this step, the time periods can be divided first to obtain drastic change periods with obvious wind speed changes and slow change periods with insignificant wind speed changes.

[0090] The change in wind speed can be characterized by the slope, and since only the drastic change is analyzed here and the direction of change is not considered, the data curve of the wind speed time series data is obtained (which can be obtained based on the least squares method, a well-known technology, and will not be elaborated on here) as the wind speed data curve, and the absolute value of the slope at each wind speed value is obtained on the wind speed data curve. The larger the absolute value of the slope, the more drastic the wind speed change.

[0091] Then all the absolute values ​​of the slopes are sorted in order to obtain an ascending sequence. In the ascending sequence, the first-order difference sequence of all the absolute values ​​of the slopes is calculated. The value in the first-order difference sequence reflects the difference between the absolute values ​​of the two slopes. When the value is larger, the gap between the two is more obvious. Therefore, the maximum value is screened out in the first-order difference sequence. The maximum value at this time represents the two absolute values ​​of the slopes with the most obvious difference in the ascending sequence, and the maximum value can divide the ascending sequence into two parts. The part before the two absolute values ​​of the slopes corresponding to the maximum value is considered to be a data set with a smaller value; and the part after the maximum value is considered to be a data set with a larger value.

[0092] Therefore, in the differential sequence, the two slope absolute values ​​corresponding to the maximum value in the ascending sequence are used as mark values, and in the ascending sequence, the wind speed value corresponding to the maximum value of the two mark values ​​and the absolute value of the slope thereafter on the wind speed data curve is used as the first wind speed value, and the wind speed value corresponding to the minimum value of the two mark values ​​and the absolute value of the slope thereafter on the wind speed data curve is used as the second wind speed value, wherein the first wind speed value is the wind speed value that will have an obvious change trend, and the second wind speed value is the wind speed value that will not have an obvious change trend.

[0093] On the wind speed data curve, the time period corresponding to the subsequence of the first wind speed values ​​that are continuous in time series is taken as the time period of rapid change, and the time period corresponding to the subsequence of the second wind speed values ​​that are continuous in time series is taken as the time period of slow change.

[0094] So far, all the time periods of rapid change and slow change are obtained.

[0095] Step S212: During the period of drastic changes, the deviation characteristics between the actual flight attitude data at each moment and the preset flight attitude data are integrated to obtain an attitude deviation factor.

[0096] In each time period of drastic changes, the absolute value of the difference between the actual flight attitude data and the preset flight attitude data at each moment is used as the attitude deviation factor. The larger the attitude deviation factor, the greater the difference between the flight attitude of the drone and the set flight attitude. Then the sum of the attitude deviation factors corresponding to all types of data values ​​is used as the attitude deviation parameter. The attitude deviation parameter integrates the deviation characteristics of all types of data values ​​in the flight attitude data. The larger the value, the greater the degree of attitude deviation.

[0097] Finally, the mean of the attitude deviation parameters at all moments in all periods of drastic changes is taken as the attitude deviation factor. The larger the attitude deviation factor, the greater the difference between the UAV and the pre-set flight state during flight. The greater the possibility of it being disturbed by the environment, the greater the degree of attitude deviation.

[0098] It should be noted that, in this embodiment of the present invention, the types of flight attitude data may include pitch angle, roll angle and heading angle, etc.; the preset flight attitude data can be obtained according to the UAV control system database.

[0099] Step S213: determining a wind speed change factor based on the difference between the wind speed values ​​in the drastic change period and the slow change period.

[0100] Calculate the mean of all wind speed values ​​during all periods of rapid change. Similarly, calculate the mean of all wind speed values ​​during all periods of slow change. The mean wind speed value here represents the average performance level of wind speed values ​​in each period. The absolute value of the difference between the two is used as the wind speed variation factor. The larger the wind speed variation factor, the more drastic the wind speed change in the environment, and the more likely the flight attitude will be affected and deviate.

[0101] Step S214: Fusing the attitude deviation factor and the wind speed change factor to determine the flight attitude deviation of the UAV.

[0102] Based on the analysis in the previous steps, it can be seen that both the attitude deviation factor and the wind speed variation factor are positively correlated with the degree of deviation of the drone's flight attitude. Therefore, the product of the attitude deviation factor and the wind speed variation factor is normalized to obtain the value of the drone's flight attitude deviation. In this case, the greater the flight attitude deviation, the greater the likelihood that the drone is subject to environmental interference. Normalization is a technical method well known to those skilled in the art. The normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited here.

[0103] The state fluctuation index and flight attitude deviation of the UAV can both reflect the impact of wind on it in a complex environment, so the combination of the two can be used to determine the UAV's wind resistance index.

[0104] Preferably, in one embodiment of the present invention, the method for obtaining the wind resistance index includes:

[0105] A larger state fluctuation index indicates a less stable drone during flight, and thus, poorer wind resistance. A larger flight attitude deviation indicates a greater likelihood of environmental interference, also indicating poorer wind resistance. Therefore, the product of the state fluctuation index and the flight attitude deviation is negatively correlated and normalized to achieve logical relationship correction, thereby obtaining the drone's wind resistance index. A larger wind resistance index indicates greater flight stability and stronger wind resistance in complex environments. Conversely, a smaller wind resistance index indicates less stable flight and greater susceptibility to environmental interference, thus requiring greater gimbal maintenance. The negative correlation mapping and normalization here can be performed using the formula exp(-x), where exp() represents an exponential function with the natural constant e as its base, and x represents the independent variable.

[0106] Step S3: Based on the variation characteristics of the actual flight attitude data of the UAV, the blur variation between the image data is analyzed, and the compensation effect index of the gimbal motor is determined in combination with the fluctuation characteristics of the output power and speed of the gimbal motor.

[0107] Amidst external interference, the gimbal must maintain drone stability and ensure image and video quality. The controllability of the micromotors directly impacts the gimbal's stabilization. The compensation effect of the gimbal's micromotors can significantly improve the accuracy of a drone's attitude adjustments during flight. Analyzing the compensation effect of these micromotors can help optimize energy management, adjust motor operating modes and power output, and ensure drone stability.

[0108] The function of the micro motor in the gimbal is to enable the gimbal to better offset the unstable changes in the drone's flight attitude during flight and produce appropriate jitter characteristics. Therefore, a better compensation effect is mainly reflected in the stable output of the micro motor's output power and speed. If fluctuations occur frequently, it can be regarded as a poor compensation effect, and the energy loss will also be large. At the same time, the compensation effect can be more directly reflected in the quality of the image data taken by the drone. Therefore, based on the changing characteristics of the drone's actual flight attitude data, the blur changes between the image data are analyzed, and combined with the fluctuation characteristics of the gimbal motor's output power and speed, the compensation effect index of the gimbal motor is determined.

[0109] Preferably, in one embodiment of the present invention, the method for obtaining the compensation effect index includes:

[0110] See also Figure 4 , which shows a flow chart of a method for obtaining a compensation effect indicator in one embodiment of the present invention, the method comprising the following steps:

[0111] Step S301: Compare the deviations between the actual flight attitude data of the UAV at any two adjacent moments, thereby screening all the change moment groups.

[0112] Between any two adjacent moments, the actual flight attitude data of the UAV is compared under all types of data values. If there is a set of different data values, the two adjacent moments are regarded as a set of change moment groups, that is, the flight attitude of the UAV has changed between these two adjacent moments, and it is very likely that the compensation effect will be poor.

[0113] Step S302: Fusion analysis is performed on the blur changes between two image data in all the time-varying groups to determine a first loss factor.

[0114] When the gimbal compensation effect is poor, the image captured by the drone will be more blurred, and the gradient changes of the pixels will be more complex. Therefore, in each change moment group, the image data at each moment is processed based on the Canny operator to obtain the gradient amplitude of the pixel point, and the variance of the gradient amplitude of all pixels is used as the blur amount. The larger the blur amount, the more complex the gradient changes of the pixels in the image.

[0115] The absolute value of the difference in the blur amount of the image data at two moments is used as the blur difference factor. The larger the blur difference factor is, the greater the impact of the change in the UAV's flight attitude on the image quality is. It can be regarded as that when the flight attitude changes, the compensation effect of the gimbal is poor. The blur amount of the image data of the previous moment in the time series is negatively correlated and mapped. The smaller the blur amount is, the better the image quality is. At this time, the larger the value after the negative correlation mapping of the blur amount is, the better the image quality is before the flight attitude changes, and the worse the compensation effect is when the flight attitude changes. Therefore, the product of the value after the negative correlation mapping of the blur amount and the blur difference factor is used as the blur eigenvalue. The larger the blur eigenvalue is, the greater the jitter effect is and the worse the compensation effect is. The negative correlation mapping here can be performed using the formula exp(-x) or Here, exp() represents an exponential function with the natural constant e as the base, and x represents the independent variable.

[0116] The first loss factor is normalized by taking the mean of the fuzzy eigenvalues ​​of all groups at the time of change. A larger first loss factor indicates that the image data captured by the drone was of good quality before the flight attitude data changed, but that the image data quality deteriorated after the flight attitude data changed, indicating a poorer compensation effect. Normalization is a well-known technique for those skilled in the art, and the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0117] It should be noted that obtaining the gradient amplitude of a pixel point by using the Canny operator is a well-known technique, and the specific process will not be described in detail here.

[0118] Step S303: The fluctuation characteristics of the output power time series data and the fluctuation characteristics of the speed time series data are integrated to obtain a second loss factor.

[0119] The variance of the output power values ​​in the output power time series data is multiplied by the variance of the speed values ​​in the speed time series data. The resulting product is normalized and used as the second loss factor. A larger variance indicates greater data fluctuation. Greater fluctuation indicates a more unstable UAV flight state, poorer gimbal compensation, and greater energy loss. Normalization is a well-known technical method for those skilled in the art. The normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited here.

[0120] Step S304: The first loss factor and the second loss factor are combined to obtain a compensation effect index of the pan / tilt motor.

[0121] Given that both the first loss factor and the second loss factor are negatively correlated with the compensation effect, in this embodiment of the present invention, the value after negative correlation mapping of the product of the first loss factor and the second loss factor is used as the compensation effect index of the gimbal motor. At this time, the smaller the compensation effect index is, the worse the compensation effect of the stabilization micromotor in the gimbal for the change of the drone's flight state during the flight of the drone. The gimbal cannot effectively offset the impact of the change in the drone's flight attitude, further increasing the energy loss of the micromotor, which is not conducive to the intelligent control of the micromotor. The negative correlation mapping and normalization here can be performed using the formula exp(-x) or Here, exp() represents an exponential function with the natural constant e as the base, and x represents the independent variable.

[0122] Step S4: adjusting the preset PID parameters based on the compensation effect index of the gimbal motor and the wind resistance index of the UAV to control the output power of the gimbal motor.

[0123] When using a PID controller to control the output power of the gimbal motor, if the preset PID parameters are used, the response time for eliminating the error will be longer when facing the rapid movement of the drone and external interference, resulting in the gimbal being unable to adjust in time, causing image jitter. The compensation effect index of the gimbal motor and the wind resistance index of the drone obtained in the above steps can respectively evaluate the gimbal's compensation for the drone's jitter and the drone's wind resistance in complex environments. Therefore, these two indicators can be integrated to adjust the preset PID parameters, thereby controlling the output power of the gimbal motor, so that it can correct the jitter of the drone during flight more promptly, ensure the stability of the flight attitude, and improve the quality of image data.

[0124] In this embodiment of the present invention, the preset PID parameters include an initial proportional gain coefficient, an initial integral gain coefficient, and an initial differential gain coefficient, all of which can be obtained based on the Ziegler-Nichols method. The specific process is a well-known technology and will not be described here.

[0125] Preferably, in one embodiment of the present invention, the preset PID parameters are adjusted based on the compensation effect index of the pan-tilt motor and the wind resistance index of the drone to control the output power of the pan-tilt motor, including:

[0126] In this embodiment of the present invention, the initial proportional gain coefficient in the preset PID parameters is mainly adjusted in order to adjust the response speed.

[0127] The smaller the compensation effect index, the worse the current gimbal's compensation effect on the drone's jitter, and the response speed needs to be improved. Therefore, the compensation effect index is negatively correlated and normalized to achieve logical relationship correction, resulting in the first adjustment factor. The larger the first adjustment factor, the greater the degree to which the response speed needs to be increased.

[0128] Similarly, given that the smaller the UAV's wind resistance index is, the more susceptible it is to the environment, and the response speed also needs to be improved, the UAV's wind resistance index is negatively correlated and normalized to achieve logical relationship correction and obtain the second adjustment factor. The larger the second adjustment factor is, the greater the degree to which the response speed needs to be increased, so as to facilitate timely error correction.

[0129] The negative correlation mapping and normalization in this process can both use the formula exp(-x), where exp() represents an exponential function with the natural constant e as the base, and x represents the independent variable.

[0130] Then the sum of the first adjustment factor and the second adjustment factor is used as the adjustment parameter, and the adjustment parameter is multiplied by the initial proportional gain coefficient of the PID controller to obtain the adjustment proportional gain coefficient. Based on the above analysis, it can be seen that the larger the adjustment proportional gain coefficient, the faster the error correction speed after adjustment, the more timely the response, the better it can ensure the flight stability of the UAV, and enhance the dynamic compensation capability.

[0131] Therefore, the output power of the pan-tilt motor is finally controlled by PID based on the adjustment of the proportional gain coefficient, the initial integral gain coefficient and the initial differential gain coefficient.

[0132] In summary, by synchronously collecting the UAV's operating parameters (angular velocity, acceleration), flight attitude data, image data, ambient wind speed, and gimbal motor output power and speed time series data, we can fully perceive the flight environment and load status, providing a data foundation for subsequent precise control. In complex environments, the preset PID parameters can result in excessively long response times, making it impossible to adjust errors in a timely manner. Therefore, we can analyze the extent of the environmental impact on the UAV and adjust the preset PID parameters based on the UAV's operating status. First, we calculate a state fluctuation index based on the difference between angular velocity trends and acceleration fluctuations. This accurately characterizes the degree of UAV vibration and attitude disturbance. This state fluctuation index provides a basis for dynamic compensation in gimbal motor control. Combining wind speed changes with attitude deviation data quantifies flight attitude deviation, reflecting the impact of environmental disturbances on UAV stability in real time. Determining flight attitude deviation allows subsequent gimbal motor control strategies to better adapt to wind disturbance trends. By integrating state fluctuations and attitude deviation, we generate a wind resistance index, dynamically assessing the UAV's stability in complex wind fields. Furthermore, based on the changing characteristics of actual flight attitude data, the effectiveness of micromotor jitter compensation can be analyzed by combining image blur changes with motor power / speed fluctuations to quantify the compensation effect. Finally, PID parameters are optimized based on wind resistance and compensation effect indicators to achieve scenario-specific adaptation of the micromotor control strategy: the compensation effect indicator constrains the output power of the gimbal motor, reducing average power consumption while ensuring gimbal stability; the wind resistance indicator enables the drone gimbal control strategy to adapt to complex environments. This allows for rapid error correction, ensuring drone stability while reducing motor energy waste.

[0133] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0134] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. An intelligent control method for a micro motor for stabilizing a UAV gimbal, characterized in that: The method comprises: Obtaining the time series data of the UAV's operating parameters, actual flight attitude data at each moment, and captured image data. The operating parameters include angular velocity and acceleration in each direction; obtaining the time series data of the wind speed in the UAV's environment; obtaining the time series data of the output power and speed of the gimbal motor; The drone's state fluctuation index is determined based on the changing trend and fluctuation characteristics of angular velocity and the fluctuation differences between accelerations in all directions. The drone's flight attitude deviation is determined based on the changing trend and numerical differences of wind speed values, as well as the deviation between the actual flight attitude data at each moment and the preset flight attitude data. The drone's wind resistance index is determined based on the flight attitude deviation and state fluctuation index. Based on the variation characteristics of the actual flight attitude data of the UAV, the fuzzy variation between the image data is analyzed. Combined with the fluctuation characteristics of the output power and speed of the gimbal motor, the compensation effect index of the gimbal motor is determined. The preset PID parameters are adjusted based on the compensation effect index of the gimbal motor and the wind resistance index of the drone to control the output power of the gimbal motor.

2. The intelligent control method for a drone gimbal stabilization micromotor according to claim 1, characterized in that: The method for obtaining the state fluctuation index includes: In the angular velocity time series data, analyze the changing trend and fluctuation characteristics of the angular velocity to determine the rapid swing index of the drone; In the acceleration time series data in each direction, the variance of all accelerations is used as the change fluctuation factor corresponding to each direction; Combine all directions in pairs to obtain all direction combinations. In each direction combination, calculate the absolute value of the difference between the change fluctuation factors corresponding to the two directions as the state change factor. Normalize the mean of the state change factors of all direction combinations to obtain the value as the state fluctuation parameter of the drone. The product of the rapid swing index and the state fluctuation parameter is normalized to a value which is used as the state fluctuation index of the UAV.

3. The intelligent control method for the UAV gimbal stabilization micro motor according to claim 2, characterized in that: The method for obtaining the fast swing index includes: Obtaining a data curve corresponding to the angular velocity time series data as an angular velocity data curve, and obtaining a slope value at each angular velocity on the angular velocity data curve; Calculate the absolute value of the difference between the slope values ​​of each two adjacent angular velocities as the trend change factor, and take the average of all trend change factors as the trend change parameter; Obtain all extreme points on the angular velocity data curve; The time interval between each two adjacent extreme points is taken as the interval factor, and the mean of all interval factors is taken as the time interval parameter; The value after normalizing the ratio of the trend change parameter to the time interval parameter is used as the rapid swing index of the UAV.

4. The intelligent control method for a drone gimbal stabilization micromotor according to claim 1, characterized in that: The method for obtaining the flight attitude deviation comprises: In the wind speed time series data, the time series is divided based on the change trend and numerical difference of the wind speed value to obtain the time period of rapid change and the time period of slow change; In each of the drastic change time periods, between the actual flight attitude data and the preset flight attitude data at each moment, the absolute value of the difference between the same type of data values ​​is used as the attitude deviation factor, and the sum of the attitude deviation factors corresponding to all types of data values ​​is used as the attitude deviation parameter; The mean of the attitude deviation parameters at all moments in all drastic change periods is taken as the attitude deviation factor; Calculating the absolute value of the difference between the mean of all wind speed values ​​in all the drastic change time periods and the mean of all wind speed values ​​in all the slow change time periods as the wind speed change factor; The product of the attitude deviation factor and the wind speed change factor is normalized to a value which is used as the flight attitude deviation degree of the UAV.

5. The intelligent control method for the UAV gimbal stabilization micro motor according to claim 4, characterized in that: In the wind speed time series data, the time series is divided based on the change trend and numerical difference of the wind speed value to obtain the rapid change time period and the slow change time period, including: Obtaining a data curve of the wind speed time series data as a wind speed data curve, and obtaining an absolute value of a slope at each wind speed value on the wind speed data curve; Sort all the absolute values ​​of the slopes in order to obtain an ascending sequence; In the ascending sequence, a difference sequence of all slope absolute values ​​is calculated, and in the difference sequence, two slope absolute values ​​corresponding to the maximum value in the ascending sequence are used as mark values; In the ascending sequence, the wind speed value corresponding to the maximum value of the two marked values ​​and the absolute value of the slope thereafter on the wind speed data curve is taken as the first wind speed value, and the wind speed value corresponding to the minimum value of the two marked values ​​and the absolute value of the slope thereafter on the wind speed data curve is taken as the second wind speed value; On the wind speed data curve, a time period corresponding to a subsequence of first wind speed values ​​that are continuous in time series is used as a time period of rapid change, and a time period corresponding to a subsequence of second wind speed values ​​that are continuous in time series is used as a time period of slow change. Get all the time periods of rapid change and slow change.

6. The intelligent control method for a drone gimbal stabilization micromotor according to claim 1, characterized in that: The method for obtaining the wind resistance index includes: The product of the state fluctuation index and the flight attitude deviation is negatively correlated and normalized to a value obtained as the wind resistance index of the UAV.

7. The intelligent control method for a drone gimbal stabilization micromotor according to claim 1, characterized in that: The method for obtaining the compensation effect index includes: Between any two adjacent moments, compare the data values ​​of the actual flight attitude data of the UAV under the same category. If there is a group of data values ​​that are different, then the two adjacent moments are regarded as a group of change moments. Fusion analysis of the blur changes between the two image data in all time change groups to determine the first loss factor; Multiplying the variance of the output power values ​​in the output power time series data by the variance of the speed values ​​in the speed time series data, and normalizing the obtained product as a second loss factor; The value obtained by performing negative correlation mapping on the product of the first loss factor and the second loss factor is used as the compensation effect indicator of the pan / tilt motor.

8. The intelligent control method for the UAV gimbal stabilization micro motor according to claim 7, characterized in that: The method for obtaining the first loss factor includes: In each group of changing moments, the image data at each moment is processed based on the Canny operator to obtain the gradient amplitude of the pixel points. The variance of the gradient amplitudes of all pixel points is used as the blur amount. The absolute value of the difference between the blur amounts of the image data at two moments is used as the blur difference factor. The product of the value after negative correlation mapping of the blur amount of the image data at the previous moment in the temporal sequence and the blur difference factor is used as the blur eigenvalue. The normalized value of the mean of the fuzzy eigenvalues ​​of all change moment groups is used as the first loss factor.

9. The intelligent control method for a drone gimbal stabilization micromotor according to claim 1, characterized in that: The method for obtaining the preset PID parameters includes: The preset PID parameters include an initial proportional gain coefficient, an initial integral gain coefficient, and an initial differential gain coefficient; The initial proportional gain coefficient, initial integral gain coefficient, and initial differential gain coefficient are obtained based on the Ziegler-Nichols method.

10. The intelligent control method for the UAV gimbal stabilization micro motor according to claim 9, characterized in that: The method of adjusting the preset PID parameters based on the compensation effect index of the pan-tilt motor and the wind resistance index of the UAV to control the output power of the pan-tilt motor includes: The compensation effect index is negatively correlated and normalized, and the result is used as a first adjustment factor; The wind resistance index of the UAV is negatively correlated and normalized, and the value obtained is used as the second adjustment factor; Taking the sum of the first adjustment factor and the second adjustment factor as an adjustment parameter; Multiplying the adjustment parameter by the initial proportional gain coefficient of the PID controller to obtain an adjustment proportional gain coefficient; The output power of the motor is controlled by PID based on the adjusted proportional gain coefficient, the initial integral gain coefficient and the initial differential gain coefficient.

Images