A multi-motor load association heavy-load unmanned aerial vehicle electric governor cooperative balancing method

Abstract

The application discloses a multi-motor load correlation heavy-load unmanned aerial vehicle electric governor cooperative balancing method, relates to the technical field of unmanned aerial vehicle electric propulsion control, and comprises the following steps: collecting motor data and flight control side instruction data, and constructing a multi-motor operation state set after preprocessing; based on the set, calculating a comprehensive load factor, combining multi-parameter correlation analysis, identifying load offset motors and cooperative compensation motors, and determining load type labels; according to the load excess of the load offset motors and the load margin of the cooperative compensation motors, generating a power transfer amount, calculating a drive duty cycle correction coefficient and a compensation coefficient; mapping the correction coefficient and the compensation coefficient into a duty cycle adjustment amount, and outputting the duty cycle adjustment amount to each electric governor after continuous smooth adjustment by an exponential smooth transition function. The load type is identified by the comprehensive load factor and the correlation analysis, the load balancing with constant total thrust is realized by combining power transfer and dynamic adjustment, and the output is continuously and smoothly adjusted, so that the thrust consistency and flight safety are improved.

Description

Technical Field

[0001] This invention relates to the field of electric propulsion control technology for unmanned aerial vehicles (UAVs), specifically a method for coordinated balancing of electric motors in heavy-load UAVs with multi-motor load association. Background Technology

[0002] Multi-rotor heavy-duty UAVs typically employ a multi-motor cooperative drive system, where multiple motors and corresponding electronic speed controllers (ESCs) work together to provide lift and attitude control torque. During actual flight, factors such as payload installation position offsets, aerodynamic disturbances, and changes in aircraft attitude can lead to differences in the load borne by each motor, resulting in uneven load distribution within the multi-motor system.

[0003] In existing technologies, flight control systems primarily distribute thrust to each motor based on attitude control requirements, achieving attitude stability by adjusting the drive duty cycle of each electronic speed controller (ESC). However, this distribution method mainly relies on feedforward control logic and lacks real-time perception and coordinated adjustment capabilities regarding the actual load status of each motor and the operating status of the ESC. When load bias or external disturbances occur, some motors become overloaded, causing the corresponding ESCs to operate at high current and high temperature for extended periods, easily triggering over-temperature protection or leading to equipment damage; while other motors are underloaded and fail to effectively utilize their thrust output capabilities. This situation not only reduces the thrust consistency of the multi-motor system but also adversely affects the overall flight safety of the aircraft.

[0004] Therefore, how to achieve load balancing and coordinated stable output in a multi-motor propulsion system has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a method for coordinated balancing of the electronic speed controllers (ESCs) of heavy-duty UAVs with multi-motor load association, so as to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for coordinated equalization of electronic speed controllers (ESCs) for heavy-load unmanned aerial vehicles (UAVs) with multi-motor load association, comprising:

[0007] S1: Collect the corresponding ESC operating data of each motor, the attitude control command data of the flight control side, and the total throttle command, perform time synchronization, outlier removal and amplitude limiting processing, and construct the operating status set of each motor. The ESC operating data includes at least bus voltage, current, speed, drive duty cycle and temperature information.

[0008] S2: Based on the set of motor states, calculate the comprehensive load factor of each motor according to the relationship between motor input power and speed and the temperature change trend. Combine the load change trend and the relationship between attitude control command change to perform correlation analysis, identify load offset motors and cooperative compensation motors, and determine the load type label, which includes: transient load or continuous load.

[0009] S3: Based on the distribution and load type labels of the load offset motor and the cooperative compensation motor, the power transfer amount between multiple motors is generated according to the load excess of the load offset motor and the load margin of the cooperative compensation motor. The drive duty cycle correction coefficient of the load offset motor and the drive duty cycle compensation coefficient of the cooperative compensation motor are calculated.

[0010] S4: Map the correction coefficient and compensation coefficient to the drive duty cycle adjustment amount of the ESC, continuously and smoothly adjust the adjusted duty cycle through an exponential smooth transition function, and output the adjusted control signal to each ESC.

[0011] The present invention is further configured such that S1 includes:

[0012] Obtain bus voltage, current, speed, drive duty cycle and temperature information sampled by the internal sensors of the corresponding ESC for each motor, and construct the ESC operating data corresponding to each motor.

[0013] Acquire attitude control commands and total throttle commands output from the flight control side;

[0014] The acquired ESC operating data and flight control side data are processed synchronously at the same time, and the data timestamps are unified.

[0015] Perform outlier removal on the time-synchronized data to filter out communication errors and data jumps;

[0016] After removing outliers, the data is subjected to amplitude limiting to ensure that each data item is within a preset physical reasonable range;

[0017] Based on the data after amplitude limiting, a first subset of states corresponding to the ESC operating data of each motor is constructed, and a second subset of states for flight control side data shared by each motor is constructed. The first subset of states and the second subset of states together constitute the operating state set of multiple motors.

[0018] The present invention is further configured such that S2 includes: a comprehensive load factor calculation and motor identification step, and a load trend analysis and type discrimination step.

[0019] The present invention is further configured such that the comprehensive load factor calculation and motor identification steps include:

[0020] The input power of each motor is calculated based on the bus voltage and current in the set of motor operating states, and the temperature rise rate of each motor is calculated based on temperature information, which is then used for subsequent load change trend analysis.

[0021] The physical load component is determined based on the ratio of input power to speed of each motor, the risk weighting coefficient is determined based on the temperature rise rate, and the physical load component is weighted and corrected to generate a comprehensive load factor.

[0022] The average load of all motors is calculated based on the comprehensive load factor of each motor.

[0023] The comprehensive load factor is compared with the average load, and motors whose comprehensive load factor exceeds the first preset multiple of the average load are identified as load-off motors.

[0024] Motors whose comprehensive load factor is lower than the second preset multiple of the average load and whose temperature is lower than the preset safety threshold are identified as collaborative compensation motors.

[0025] The present invention is further configured such that the load trend analysis and type determination step includes:

[0026] Based on the relationship between the comprehensive load factor of each motor and time, the load change trend of each motor is calculated.

[0027] Calculate the rate of change of at least one attitude control command based on the attitude control commands from the flight control side;

[0028] The correlation coefficient between the load change trend of each motor and the change rate of attitude control command was calculated using the sliding window correlation analysis method.

[0029] When the correlation coefficient is greater than the preset correlation threshold, the load type of the corresponding motor is identified as a transient load coupled with attitude.

[0030] When the correlation coefficient is not greater than the preset correlation threshold, the load type of the corresponding motor is identified as a continuous load in order to avoid incorrect adjustment of the transient load caused by attitude control.

[0031] The present invention is further configured such that S3 includes: a power redistribution calculation step, a basic correction coefficient generation step, and a dynamic adjustment and coefficient correction step.

[0032] The present invention is further configured such that the power redistribution calculation step includes:

[0033] Obtain the difference between the comprehensive load factor of each load offset motor and the average load, perform an accumulation operation on the difference of each load offset motor, and determine the accumulation result as the load exceeding the total amount;

[0034] Obtain the difference between the average load and the comprehensive load factor of each collaboratively compensated motor, perform an accumulation operation on the difference of each collaboratively compensated motor, and determine the accumulation result as the total load margin.

[0035] The load exceeding the total amount is compared with the total load margin, and the smaller of the two values ​​is selected as the power transfer amount for load redistribution.

[0036] The present invention is further configured such that the basic correction coefficient generation step includes:

[0037] For load offset motors, a proportional allocation method based on the difference ratio is adopted. According to the proportion of the difference between the comprehensive load factor and the average load of each load offset motor in the total load excess, the drive duty cycle correction coefficient corresponding to each load offset motor is generated.

[0038] For the collaborative compensation motors, a proportional allocation method based on the difference ratio is adopted. According to the proportion of the difference between the average load and the comprehensive load factor of each collaborative compensation motor in the total load margin, the drive duty cycle compensation coefficient corresponding to each collaborative compensation motor is generated.

[0039] The generated correction and compensation coefficients are normalized based on the power transfer amount to ensure that the power reduction of the load offset motor is equal to the power increase of the cooperative compensation motor, thus satisfying the constraint that the overall output of the multi-motor system remains consistent.

[0040] The present invention is further configured such that the dynamic adjustment and coefficient correction step includes:

[0041] For load-off motors, the drive duty cycle correction coefficient is adjusted in segments according to the load change trend of each motor. Different adjustment ranges are used when the load change trend is within the preset value range. The greater the load change trend, the greater the adjustment range.

[0042] For the collaborative compensation motor, the drive duty cycle compensation coefficient is adjusted differently based on the load type label obtained from the load trend analysis and type discrimination steps. When the load type label is attitude-coupled transient load, the compensation strength of the compensation coefficient is reduced, and when the load type label is continuous load, the compensation strength of the compensation coefficient is increased.

[0043] The output is the final correction coefficient of the load offset motor and the final compensation coefficient of the collaborative compensation motor after adjustment.

[0044] The present invention is further configured such that S4 includes:

[0045] The final correction coefficient of the load offset motor and the final compensation coefficient of the cooperative compensation motor are mapped to the drive duty cycle of the corresponding ESC, respectively, to generate the drive duty cycle adjustment amount of each ESC.

[0046] An exponential smooth transition function is used to continuously and smoothly adjust the drive duty cycle of each ESC to generate a smooth transition drive duty cycle.

[0047] The smoothly transitioned drive duty cycle is output as a control signal to each corresponding ESC.

[0048] This invention provides a method for coordinated load balancing of ESCs in heavy-load UAVs with multi-motor load association. The method involves: S1: collecting operating data of the corresponding ESCs for each motor, attitude control command data from the flight controller, and total throttle command; performing time synchronization, outlier removal, and amplitude limiting processing to construct a set of operating states for each motor. The ESC operating data includes at least bus voltage, current, speed, drive duty cycle, and temperature information. S2: based on the motor state set, calculating the comprehensive load factor of each motor according to the relationship between motor input power and speed, combined with temperature change trends, and performing correlation analysis based on load change trends and attitude control command changes to identify load-shifted motors and collaboratively compensated motors. S3: Based on the distribution of the load offset motor and the cooperative compensation motor and the load type label, according to the load excess of the load offset motor and the load margin of the cooperative compensation motor, the power transfer amount between multiple motors is generated, and the drive duty cycle correction coefficient of the load offset motor and the drive duty cycle compensation coefficient of the cooperative compensation motor are calculated; S4: The correction coefficient and the compensation coefficient are mapped to the drive duty cycle adjustment amount of the ESC, and the adjusted duty cycle is continuously and smoothly adjusted through an exponential smooth transition function, and the adjusted control signal is output to each ESC. The beneficial effects include:

[0049] By constructing a comprehensive load factor that integrates temperature rise trend weighting and combining the correlation analysis between load change trend and attitude command change rate, it is possible to accurately identify load offset motors and collaborative compensation motors, and distinguish between transient disturbance loads and continuous off-center loads, providing accurate decision-making basis for subsequent differentiated equalization control.

[0050] Based on the power transfer amount generated by the load excess and load margin, the correction coefficient and compensation coefficient are generated by combining proportional allocation and normalization processing. This ensures that the total thrust of the multi-motor system remains unchanged while realizing the dynamic redistribution of the load. At the same time, the correction coefficient is adjusted in segments and compensated differently according to the load change trend and load type, thus realizing adaptive and conflict-free collaborative balance control.

[0051] By using an exponential smooth transition function to continuously and smoothly adjust the duty cycle, current surges and flight control oscillations caused by sudden changes in control commands are avoided. Combined with real-time sensing of temperature rise trends and closed-loop iterative correction, the risk of local overload is effectively reduced, and the thrust consistency, control stability, and overall flight safety of the multi-motor system are improved.

[0052] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0054] Figure 1 The flowchart illustrates a method for collaborative equalization of ESCs in heavy-duty UAVs with multi-motor load association, as an exemplary embodiment of the present invention. Detailed Implementation

[0055] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0056] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0057] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0058] Example:

[0059] A collaborative equalization method for ESCs of heavy-duty UAVs with multi-motor load association, such as Figure 1 As shown, it includes:

[0060] S1: Collect the corresponding ESC operating data of each motor, the attitude control command data of the flight control side, and the total throttle command, perform time synchronization, outlier removal and amplitude limiting processing, and construct the operating status set of each motor. The ESC operating data includes at least bus voltage, current, speed, drive duty cycle and temperature information.

[0061] S2: Based on the set of motor states, calculate the comprehensive load factor of each motor according to the relationship between motor input power and speed and the temperature change trend. Combine the load change trend and the relationship between attitude control command change to perform correlation analysis, identify load offset motors and cooperative compensation motors, and determine the load type label, which includes: transient load or continuous load.

[0062] S3: Based on the distribution and load type labels of the load offset motor and the cooperative compensation motor, the power transfer amount between multiple motors is generated according to the load excess of the load offset motor and the load margin of the cooperative compensation motor. The drive duty cycle correction coefficient of the load offset motor and the drive duty cycle compensation coefficient of the cooperative compensation motor are calculated.

[0063] S4: Map the correction coefficient and compensation coefficient to the drive duty cycle adjustment amount of the ESC, continuously and smoothly adjust the adjusted duty cycle through an exponential smooth transition function, and output the adjusted control signal to each ESC.

[0064] The present invention is further configured such that S1 includes:

[0065] Obtain bus voltage, current, speed, drive duty cycle and temperature information sampled by the internal sensors of the corresponding ESC for each motor, and construct the ESC operating data corresponding to each motor.

[0066] Acquire attitude control commands and total throttle commands output from the flight control side;

[0067] The acquired ESC operating data and flight control side data are processed synchronously at the same time, and the data timestamps are unified.

[0068] Perform outlier removal on the time-synchronized data to filter out communication errors and data jumps;

[0069] After removing outliers, the data is subjected to amplitude limiting to ensure that each data item is within a preset physical reasonable range;

[0070] Based on the data after amplitude limiting, a first subset of states corresponding to the ESC operating data of each motor is constructed, and a second subset of states for flight control side data shared by each motor is constructed. The first subset of states and the second subset of states together constitute the operating state set of multiple motors. Specifically, this embodiment uses a quadcopter heavy-duty UAV as an example. This UAV is equipped with four motors and four corresponding electronic speed controllers. To achieve multi-motor load balancing control, it is first necessary to construct a set of operating states for each motor. The specific implementation process is as follows: After the collaborative balancing control module is started, it sends a data request command to the four electronic speed controllers through the controller local area network bus. After receiving the request, each electronic speed controller reads the sampled values ​​of its internal sensors at a sampling frequency of 100 Hz, and obtains five parameters: bus voltage, current, speed, drive duty cycle, and temperature. The unit of bus voltage is volts, the unit of current is amperes, the unit of speed is revolutions per minute, the unit of drive duty cycle is a percentage, and the unit of temperature is degrees Celsius. Each electronic speed controller packages the above five parameters together with its local timestamp into a data frame and sends it back to the collaborative balancing control module. At the same time, the collaborative balancing control module reads attitude control commands and total throttle commands from the flight controller through the serial peripheral interface. The attitude control commands include roll angle commands, pitch angle commands, and yaw angle commands, all in degrees. The total throttle command is expressed as a percentage, and the receiving time is recorded. After receiving all data, the Cooperative Equalization Control Module uses the timestamp of the flight controller data as a reference, traversing the data frames of the four electronic speed controllers. It compares the timestamp of each electronic speed controller's data with the timestamp of the flight controller's data. For data with a time deviation exceeding 10 milliseconds, a linear interpolation method is used for time alignment. This involves calculating the data value corresponding to the aligned time point based on the time difference and data value between two consecutive valid data points, thus eliminating time asynchrony caused by data transmission delays. When consecutive valid data points are missing or continuous data loss occurs, the most recent valid data is retained or the current cycle is skipped. After time synchronization is complete, the Cooperative Equalization Control Module performs outlier removal on the five parameters of each electronic speed controller and the three attitude command parameters of the flight controller. Taking current parameters as an example, the module maintains a sliding window containing the 10 most recent valid current values. It calculates the arithmetic mean and standard deviation of the data within the window, compares the current current value with the mean, and if the deviation is greater than three times the standard deviation, it is considered an outlier and replaced with the mean value within the window. Simultaneously, the outlier is logged. This method effectively filters out data fluctuations caused by communication interference or occasional sensor malfunctions. Other parameters are handled using the same or equivalent outlier detection method. After outlier removal, the collaborative equalization control module performs amplitude limiting on each parameter to ensure that each data item is within a preset physically reasonable range.The bus voltage limit range is 20V to 60V, corresponding to the full battery voltage plus 10% and the ESC undervoltage protection threshold; the current limit range is 0A to 120A, corresponding to 1.2 times the rated current of the ESC (Electronic Speed ​​Controller) of 100A; the speed limit range is 0rpm to 12000rpm, corresponding to the maximum no-load speed of the motor; the drive duty cycle limit range is 0% to 100%; the temperature limit range is -20°C to 85°C, where 85°C is the ESC over-temperature protection threshold minus 5°C; the roll and pitch command limits are -30°C to +30°C, corresponding to the maximum permissible attitude angles of the aircraft; the yaw command limit is -180°C to +180°C; and the total throttle command limit is 0% to 100%. When a data item exceeds a preset range, it is limited to the corresponding upper or lower limit value. These limiting ranges are preset based on the rated parameters of the motor and electronic speed controller or loaded via a configuration file. After limiting, the collaborative equalization control module stores the limited data from the four electronic speed controllers into four structure variables. Each structure contains five members: bus voltage, current, speed, drive duty cycle, and temperature, forming a first subset of states corresponding to the four motors. Simultaneously, the roll angle, pitch angle, yaw angle, and total throttle commands from the flight controller are stored in another structure variable, forming a second subset of states shared by all motors. The four first and second subsets are combined and encapsulated to form the operating state set for the current control cycle, stored in a shared memory area for subsequent load factor calculation and correlation determination.

[0071] The present invention is further configured such that S2 includes: a comprehensive load factor calculation and motor identification step, and a load trend analysis and type determination step;

[0072] The steps for calculating the comprehensive load factor and identifying the motor include:

[0073] The input power of each motor is calculated based on the bus voltage and current in the set of motor operating states, and the temperature rise rate of each motor is calculated based on temperature information, which is then used for subsequent load change trend analysis.

[0074] The physical load component is determined based on the ratio of input power to speed of each motor, the risk weighting coefficient is determined based on the temperature rise rate, and the physical load component is weighted and corrected to generate a comprehensive load factor.

[0075] The average load of all motors is calculated based on the comprehensive load factor of each motor.

[0076] The comprehensive load factor is compared with the average load, and motors whose comprehensive load factor exceeds the first preset multiple of the average load are identified as load-off motors.

[0077] Motors with a comprehensive load factor lower than the second preset multiple of the average load and a temperature lower than the preset safety threshold are identified as motors with collaborative compensation.

[0078] The load trend analysis and type determination steps include:

[0079] Based on the relationship between the comprehensive load factor of each motor and time, the load change trend of each motor is calculated.

[0080] Calculate the rate of change of at least one attitude control command based on the attitude control commands from the flight control side;

[0081] The correlation coefficient between the load change trend of each motor and the change rate of attitude control command was calculated using the sliding window correlation analysis method.

[0082] When the correlation coefficient is greater than the preset correlation threshold, the load type of the corresponding motor is identified as a transient load coupled with attitude.

[0083] When the correlation coefficient is not greater than a preset correlation threshold, the load type of the corresponding motor is identified as a continuous load to avoid erroneous adjustment of transient load caused by attitude control. Specifically, after completing step S1 to construct the operating state set, the cooperative equalization control module reads the operating state set of the current control cycle from the shared memory. The operating state set includes a first state subset corresponding to each of the four motors and a second state subset corresponding to the flight controller. The first state subset includes bus voltage, current, speed, and temperature, while the second state subset includes roll angle command, pitch angle command, yaw angle command, and total throttle command. Step S2 is divided into two sub-steps: comprehensive load factor calculation and motor identification, and load trend analysis and type determination. The specific implementation process is as follows: First, comprehensive load factor calculation and motor identification are performed. For each of the four motors, the input power is calculated based on the bus voltage and current. The input power equals the bus voltage multiplied by the current, in watts. The temperature rise rate is calculated based on temperature information. The temperature rise rate equals the difference between the temperature value of the current control cycle and the temperature value of the previous control cycle, divided by the control cycle duration of 0.01 seconds, in degrees Celsius per second. After calculation, the temperature rise rate is limited or smoothed to eliminate sampling. Abnormal fluctuations caused by noise; the physical load component is determined based on the ratio of input power to rotational speed, where the physical load component equals input power divided by rotational speed. When the rotational speed is less than a preset minimum threshold, the minimum threshold is used to replace or skip the calculation for that cycle; a risk weighting coefficient is determined based on the temperature rise rate, with the following rules: when the temperature rise rate is less than or equal to 0 degrees Celsius per second, the risk weighting coefficient is 0; when the temperature rise rate is greater than 0 degrees Celsius per second and less than or equal to 2 degrees Celsius per second, the risk weighting coefficient is 0.5; when the temperature rise rate is greater than 2 degrees Celsius per second, the risk weighting coefficient is 1.0. The comprehensive load factor equals the physical load component multiplied by 1 (in parentheses) plus the risk weighting coefficient. After calculating the comprehensive load factor of the four motors, the average load of all motors is calculated. The average load is equal to the sum of the comprehensive load factors of the four motors divided by 4. The comprehensive load factor is compared with the average load. The first preset multiple is 1.2 times. When the comprehensive load factor of a motor is greater than the average load multiplied by 1.2, the motor is identified as a load offset motor. The second preset multiple is 0.8 times. The preset safety threshold is 80 degrees Celsius. When the comprehensive load factor of a motor is less than the average load multiplied by 0.8 and the temperature of the motor is below 80 degrees Celsius, the motor is identified as a collaborative compensation motor.Next, load trend analysis and type determination are performed. For each motor, the load change trend is calculated based on the relationship between the comprehensive load factor and time. The load change trend is equal to the difference between the comprehensive load factor of the current control cycle and the comprehensive load factor of the previous control cycle, divided by the control cycle duration of 0.01 seconds. The roll angle command change rate and pitch angle command change rate are calculated based on the attitude control commands from the flight controller. The roll angle command change rate is equal to the difference between the roll angle command of the current control cycle and the roll angle command of the previous control cycle, divided by the control cycle duration of 0.01 seconds. The pitch angle command change rate is calculated similarly. A sliding window correlation analysis method is used to calculate the correlation coefficient between the load change trend of each motor and the attitude control command change rate. The sliding window length is 50 control cycles, corresponding to a time span of 0.5 seconds. For each motor, the load change trend sequence of the past 50 control cycles is correlated with the roll angle command rate of change sequence using Pearson correlation calculation to obtain a first correlation coefficient. The load change trend sequence is then correlated with the pitch angle command rate of change sequence using Pearson correlation calculation to obtain a second correlation coefficient. The maximum value between the first and second correlation coefficients is taken as the final correlation coefficient for that motor. If the data within the sliding window is less than a preset length or the variance of the correlation data is less than a preset threshold, the correlation calculation is skipped or the correlation coefficient is set to a preset default value. The preset correlation threshold is 0.7. When the final correlation coefficient is greater than 0.7, the load type of the motor is classified as a transient load coupled with attitude; when the final correlation coefficient is not greater than 0.7, the load type of the motor is classified as a continuous load. The calculation process is illustrated using specific values ​​from one control cycle of a quadcopter drone as an example: Motor 1 has a bus voltage of 48 volts, a current of 20 amps, an input power of 960 watts, a rotation speed of 4800 revolutions per minute, a physical load component of 0.2, a current temperature of 65 degrees Celsius, a previous cycle temperature of 63 degrees Celsius, a temperature rise rate of 200 degrees Celsius per second, a risk weighting coefficient of 1.0, and a comprehensive load factor of 0.4. Motor 2 has a bus voltage of 48 volts, a current of 18 amps, an input power of 864 watts, a rotation speed of 4900 revolutions per minute, a physical load component of 0.176, a current temperature of 62 degrees Celsius, a previous cycle temperature of 61 degrees Celsius, a temperature rise rate of 100 degrees Celsius per second, a risk weighting coefficient of 1.0, and a comprehensive load factor of 0.352. The bus voltage of motor three is 48 volts, the current is 15 amps, the input power is 720 watts, the speed is 5100 rpm, the physical load component is 0.141, the current temperature is 58 degrees Celsius, the temperature of the previous cycle was 57 degrees Celsius, the temperature rise rate is 100 degrees Celsius per second, the risk weighting coefficient is 1.0, and the comprehensive load factor is 0.282.Motor 4 has a bus voltage of 48 volts, a current of 22 amps, an input power of 1056 watts, a speed of 4700 rpm, a physical load component of 0.225, a current temperature of 68 degrees Celsius, a previous cycle temperature of 66 degrees Celsius, a temperature rise rate of 200 degrees Celsius per second, a risk weighting coefficient of 1.0, and a comprehensive load factor of 0.45. The comprehensive load factors of the four motors are 0.4, 0.352, 0.282, and 0.45, respectively, with an average load of 0.371. Multiplying the average load by 1.2 gives 0.445. Motor 4's comprehensive load factor of 0.45 is greater than 0.445, therefore Motor 4 is identified as a load offset motor. Multiplying the average load by 0.8 gives 0.297. Motor 3's comprehensive load factor of 0.282 is less than 0.297, and its temperature of 58 degrees Celsius is below 80 degrees Celsius, therefore Motor 3 is identified as a collaborative compensation motor. For load trend analysis and type identification, historical data from the past 50 control cycles are read from the circular buffer. The current cycle comprehensive load factor of motor four is 0.45, the previous cycle comprehensive load factor is 0.44, and the load change trend is 1.0. The Pearson correlation coefficient between the load change trend sequence of the past 50 cycles and the roll angle command change rate sequence is 0.82, and the Pearson correlation coefficient with the pitch angle command change rate sequence is 0.15. The final correlation coefficient is 0.82, which is greater than the preset correlation threshold of 0.7. Therefore, the load type of motor four is identified as transient load coupled with attitude. The current cycle comprehensive load factor of motor 3 is 0.282, the previous cycle comprehensive load factor was 0.28, and the load change trend is 0.2. The Pearson correlation coefficient between the load change trend sequence of the past 50 cycles and the roll angle command change rate sequence is 0.12, and the Pearson correlation coefficient with the pitch angle command change rate sequence is 0.08. The final correlation coefficient is 0.12, which is not greater than 0.7. Therefore, the load type of motor 3 is determined to be continuous load. After completing the above calculations, the load offset motor set, the cooperative compensation motor set, the comprehensive load factor of each motor, the load change trend of each motor, and the load type label of each motor are output to step S3.

[0084] The present invention is further configured such that S3 includes: a power redistribution calculation step, a basic correction coefficient generation step, and a dynamic adjustment and coefficient correction step;

[0085] The power redistribution calculation steps include:

[0086] Obtain the difference between the comprehensive load factor of each load offset motor and the average load, perform an accumulation operation on the difference of each load offset motor, and determine the accumulation result as the load exceeding the total amount;

[0087] Obtain the difference between the average load and the comprehensive load factor of each collaboratively compensated motor, perform an accumulation operation on the difference of each collaboratively compensated motor, and determine the accumulation result as the total load margin.

[0088] The load exceeding the total amount is compared with the total load margin, and the smaller of the two values ​​is selected as the power transfer amount for load redistribution.

[0089] The steps for generating the basic correction coefficients include:

[0090] For load offset motors, a proportional allocation method based on the difference ratio is adopted. According to the proportion of the difference between the comprehensive load factor and the average load of each load offset motor in the total load excess, the drive duty cycle correction coefficient corresponding to each load offset motor is generated.

[0091] For the collaborative compensation motors, a proportional allocation method based on the difference ratio is adopted. According to the proportion of the difference between the average load and the comprehensive load factor of each collaborative compensation motor in the total load margin, the drive duty cycle compensation coefficient corresponding to each collaborative compensation motor is generated.

[0092] The generated correction and compensation coefficients are normalized based on the power transfer amount to ensure that the power reduction of the load offset motor is equal to the power increase of the cooperative compensation motor, thus satisfying the constraint that the overall output of the multi-motor system remains consistent.

[0093] The dynamic adjustment and coefficient correction steps include:

[0094] For load-off motors, the drive duty cycle correction coefficient is adjusted in segments according to the load change trend of each motor. Different adjustment ranges are used when the load change trend is within the preset value range. The greater the load change trend, the greater the adjustment range.

[0095] For the collaborative compensation motor, the drive duty cycle compensation coefficient is adjusted differently based on the load type label obtained from the load trend analysis and type discrimination steps. When the load type label is attitude-coupled transient load, the compensation strength of the compensation coefficient is reduced, and when the load type label is continuous load, the compensation strength of the compensation coefficient is increased.

[0096] The output is the final correction coefficient for the load offset motor and the final compensation coefficient for the collaborative compensation motor after adjustment. Specifically, step S3 is divided into three sub-steps: power redistribution calculation, basic correction coefficient generation, and dynamic adjustment and coefficient correction. The specific implementation process is as follows: The collaborative balancing control module obtains the comprehensive load factor of each motor and the system average load from step S2, and classifies each motor based on preset judgment rules. The preset judgment rules are set as follows: the first preset multiple is 1.2, the second preset multiple is 0.8, and the preset safety threshold is 80 degrees Celsius. Motors with a comprehensive load factor greater than the average load multiplied by the first preset multiple are identified as load offset motors; motors with a comprehensive load factor less than the average load multiplied by the second preset multiple and a temperature lower than the preset safety threshold are identified as collaborative compensation motors. At the same time, the load change trend and load type label corresponding to each motor are obtained. Subsequently, for each load offset motor, the difference between its comprehensive load factor and the average load is calculated. Since the comprehensive load factor of the load offset motor is greater than the average load, the difference is naturally positive. All differences are accumulated to obtain the total load excess. For motors with collaborative compensation, the difference between the average load and its comprehensive load factor is calculated for each motor. Since the comprehensive load factor of a motor with collaborative compensation is less than the average load, this difference is naturally positive. All differences are summed to obtain the total load margin. The total load excess is compared with the total load margin, and the smaller value is selected as the power transfer amount for load redistribution. Based on this, for motors with load deviation, the correction weight is determined according to the proportion of the difference in the total load excess, and the drive duty cycle correction coefficient is calculated according to the method of "1 minus (basic adjustment coefficient multiplied by correction weight)". The basic adjustment coefficient can be set as a preset constant, with a value between 0 and 1, to control the intensity of each adjustment and avoid overshoot. For example, a basic adjustment coefficient of 0.5 means that each adjustment corrects a maximum of 50% of the load deviation. For motors with collaborative compensation, the compensation weight is determined according to the proportion of the difference in the total load margin, and the drive duty cycle compensation coefficient is calculated according to the method of "1 plus (basic compensation coefficient multiplied by compensation weight)". The basic compensation coefficient can be set as a preset constant, ranging from 0 to 1, to control the intensity of each compensation and avoid overcompensation. For example, a basic compensation coefficient of 0.5 means that each compensation utilizes a maximum of 50% of the load margin. Furthermore, a power transfer amount is introduced to normalize the correction and compensation coefficients. This normalization replaces the aforementioned correction and compensation coefficients based on the basic adjustment and compensation coefficients, ensuring that the total power reduction on the load offset side is equal to the total power increase on the compensation side. The normalized correction coefficient for the load offset motor is calculated by subtracting the ratio of the power transfer amount to its comprehensive load factor from 1, while the normalized compensation coefficient for the collaborative compensation motor is calculated by adding the ratio of the power transfer amount to its comprehensive load factor to 1.After normalization, the power reduction of the load-off motor is the power transfer amount, and the power increase of the co-compensation motor is also the power transfer amount. Since they are equal, the constraint of maintaining consistent overall output in the multi-motor system is satisfied. After normalization, for the load-off motor, segmented adjustment is performed based on the absolute value of the load change trend. The preset numerical range can be divided as follows: when the absolute value of the load change trend is less than or equal to the first threshold, it belongs to the slow change range, and the adjustment amplitude coefficient is the first coefficient; when the absolute value of the load change trend is greater than the first threshold and less than or equal to the second threshold, it belongs to the moderate change range, and the adjustment amplitude coefficient is the second coefficient; when the absolute value of the load change trend is greater than the second threshold, it belongs to the rapid change range, and the adjustment amplitude coefficient is the third coefficient. The first threshold can be set to 0.5, the second threshold to 2.0, the first coefficient to 1.0, the second coefficient to 1.2, and the third coefficient to 1.5. An adjustment amplitude coefficient greater than 1 indicates increased adjustment intensity, and less than 1 indicates decreased adjustment intensity. Multiplying the normalization correction coefficient by the adjustment amplitude coefficient yields the adjusted correction coefficient. For collaboratively compensated motors, differentiated adjustments are performed based on the load type label. When the load type label is attitude-coupled transient load, the normalized compensation coefficient is attenuated by subtracting the transient attenuation coefficient from the compensation coefficient. The transient attenuation coefficient can be set to a preset constant, such as 0.2. When the load type label is continuous load, the normalized compensation coefficient is enhanced by adding a continuous enhancement coefficient to the compensation coefficient. The continuous enhancement coefficient can be set to a preset constant, such as 0.1. This yields the adjusted compensation coefficient. Finally, boundary constraints are applied to all correction and compensation coefficients. The boundary constraints are as follows: the upper limit of the correction coefficient can be set to 1.0, and the lower limit to 0.5; the upper limit of the compensation coefficient can be set to 1.5, and the lower limit to 1.0. When the adjusted coefficient exceeds the above range, it is limited to the corresponding upper or lower limit value. For motors not identified as load offset motors or collaboratively compensated motors, both the final correction and final compensation coefficients are set to 1.0, meaning the drive duty cycle remains unchanged. The final correction coefficient and final compensation coefficient corresponding to each motor are output to step S4 for updating and executing the drive signal.

[0097] The present invention is further configured such that S4 includes:

[0098] The final correction coefficient of the load offset motor and the final compensation coefficient of the cooperative compensation motor are mapped to the drive duty cycle of the corresponding ESC, respectively, to generate the drive duty cycle adjustment amount of each ESC.

[0099] An exponential smooth transition function is used to continuously and smoothly adjust the drive duty cycle of each ESC to generate a smooth transition drive duty cycle.

[0100] The smoothed drive duty cycle is output as a control signal to each corresponding ESC. Specifically, the collaborative balancing control module obtains the final coefficients of each motor from step S3. The final correction coefficients of the motors identified as having load offset, the final compensation coefficients of the collaboratively compensated motors, and the final coefficients of the remaining motors are all variable parameters, which can be calculated or simulated based on the actual system. The example values ​​below are only for illustrating the calculation process and do not limit the implementation scope. The current drive duty cycle of each ESC is read in real time by the collaborative balancing control module through the controller local area network bus. The current drive duty cycle value is also an exemplary reference; in actual use, it should be obtained based on real-time measurements. First, a mapping operation is performed, mapping the final coefficients of each motor to the drive duty cycle of the corresponding ESC, generating a drive duty cycle adjustment amount. The drive duty cycle adjustment amount is equal to the current drive duty cycle multiplied by the final coefficient. Both the final coefficient and the current drive duty cycle are values ​​obtained in real time or calculated, not fixed values. Subsequently, an exponential smoothing transition function is used to continuously and smoothly adjust the drive duty cycle adjustment amount of each ESC, generating a smoothed drive duty cycle. The formula for calculating the exponential smooth transition function: That is, approximation ratio It equals 1 minus the negative control period of the natural constant divided by the power of the time constant, where, The approximation ratio indicates the degree to which the drive duty cycle approaches the target value within each control cycle; control cycle duration. The system timer of the collaborative balancing control module is set, for example, to 0.01 seconds, but the specific value is determined based on the processor's computing power and real-time requirements; time constant. This is a preset adjustable parameter used to control the speed of the smooth transition. The value range is, for example, 0.01 seconds to 0.1 seconds. Smaller values ​​result in a faster response but may cause a shock, while larger values ​​result in a smoother transition but increased response delay. Specific values ​​can be determined through simulation or experimental calibration based on the system's dynamic response characteristics. At the beginning of each control cycle, the current actual output drive duty cycle is read, the difference between the target drive duty cycle and the current actual output is calculated, and this difference is multiplied by an approximation ratio to obtain the adjustment increment. The current actual output is then added to this increment to obtain the drive duty cycle for this cycle. This process can be repeated until the drive duty cycle approaches the target value. An error threshold is used to determine whether the drive duty cycle has approached the target value; for example, it can be set to 0.1%. When the absolute deviation between the current actual output and the target value is less than this threshold, the target value is considered to have been reached. The actual value changes in real time according to the system state and has no fixed numerical limit. Finally, the smoothed drive duty cycle is output as a control signal to each corresponding electronic speed controller. The collaborative equalization control module calculates the drive duty cycle after a smooth transition in each control cycle and sends the drive duty cycle command to the corresponding electronic speed controller via the controller area network bus. The control cycle duration is set by the system timer, exemplarily set to 0.01 seconds, and the error threshold can be adjusted according to the system accuracy requirements. After receiving the command, the electronic speed controller converts the drive duty cycle into a pulse width modulation signal and outputs it to the motor to control the motor speed and torque.

[0101] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for coordinated balancing of ESCs in heavy-load UAVs with multi-motor load association, characterized in that, include: S1: Collect the corresponding ESC operating data of each motor, the attitude control command data of the flight control side, and the total throttle command, perform time synchronization, outlier removal and amplitude limiting processing, and construct the operating status set of each motor. The ESC operating data includes at least bus voltage, current, speed, drive duty cycle and temperature information. S2: Based on the set of motor states, calculate the comprehensive load factor of each motor according to the relationship between motor input power and speed and the temperature change trend. Combine the load change trend and the relationship between attitude control command change to perform correlation analysis, identify load offset motors and cooperative compensation motors, and determine the load type label, which includes: transient load or continuous load. S3: Based on the distribution and load type labels of the load offset motor and the cooperative compensation motor, the power transfer amount between multiple motors is generated according to the load excess of the load offset motor and the load margin of the cooperative compensation motor. The drive duty cycle correction coefficient of the load offset motor and the drive duty cycle compensation coefficient of the cooperative compensation motor are calculated. S4: Map the correction coefficient and compensation coefficient to the drive duty cycle adjustment amount of the ESC, continuously and smoothly adjust the adjusted duty cycle through an exponential smooth transition function, and output the adjusted control signal to each ESC.

2. The method for coordinated balancing of multi-motor load-associated heavy-load UAV ESCs according to claim 1, characterized in that, S1 includes: Obtain bus voltage, current, speed, drive duty cycle and temperature information sampled by the internal sensors of the corresponding ESC for each motor, and construct the ESC operating data corresponding to each motor. Acquire attitude control commands and total throttle commands output from the flight control side; The acquired ESC operating data and flight control side data are processed synchronously at the same time, and the data timestamps are unified. Perform outlier removal on the time-synchronized data to filter out communication errors and data jumps; After removing outliers, the data is subjected to amplitude limiting to ensure that each data item is within a preset physical reasonable range; Based on the data after amplitude limiting, a first subset of states corresponding to the ESC operating data of each motor is constructed, and a second subset of states for flight control side data shared by each motor is constructed. The first subset of states and the second subset of states together constitute the operating state set of multiple motors.

3. The method for coordinated balancing of multi-motor load-associated heavy-load UAV ESCs according to claim 1, characterized in that, The S2 includes: a comprehensive load factor calculation and motor identification step, and a load trend analysis and type determination step.

4. The method for coordinated balancing of multi-motor load association in heavy-load UAVs according to claim 3, characterized in that, The steps for calculating the comprehensive load factor and identifying the motor include: The input power of each motor is calculated based on the bus voltage and current in the set of motor operating states, and the temperature rise rate of each motor is calculated based on temperature information, which is then used for subsequent load change trend analysis. The physical load component is determined based on the ratio of input power to speed of each motor, the risk weighting coefficient is determined based on the temperature rise rate, and the physical load component is weighted and corrected to generate a comprehensive load factor. The average load of all motors is calculated based on the comprehensive load factor of each motor. The comprehensive load factor is compared with the average load, and motors whose comprehensive load factor exceeds the first preset multiple of the average load are identified as load-off motors. Motors whose comprehensive load factor is lower than the second preset multiple of the average load and whose temperature is lower than the preset safety threshold are identified as collaborative compensation motors.

5. The multi-motor load association method for heavy-load UAV ESC cooperative equalization according to claim 4, characterized in that, The load trend analysis and type determination steps include: Based on the relationship between the comprehensive load factor of each motor and time, the load change trend of each motor is calculated. Calculate the rate of change of at least one attitude control command based on the attitude control commands from the flight control side; The correlation coefficient between the load change trend of each motor and the change rate of attitude control command was calculated using the sliding window correlation analysis method. When the correlation coefficient is greater than the preset correlation threshold, the load type of the corresponding motor is identified as a transient load coupled with attitude. When the correlation coefficient is not greater than the preset correlation threshold, the load type of the corresponding motor is identified as a continuous load in order to avoid incorrect adjustment of the transient load caused by attitude control.

6. The method for coordinated balancing of ESCs in heavy-load UAVs with multi-motor load association according to claim 1, characterized in that, The S3 includes: a power redistribution calculation step, a basic correction coefficient generation step, and a dynamic adjustment and coefficient correction step.

7. The method for coordinated balancing of multi-motor load association in heavy-load UAV ESCs according to claim 6, characterized in that, The power redistribution calculation steps include: Obtain the difference between the comprehensive load factor of each load offset motor and the average load, perform an accumulation operation on the difference of each load offset motor, and determine the accumulation result as the load exceeding the total amount; Obtain the difference between the average load and the comprehensive load factor of each collaboratively compensated motor, perform an accumulation operation on the difference of each collaboratively compensated motor, and determine the accumulation result as the total load margin. The load exceeding the total amount is compared with the total load margin, and the smaller of the two values ​​is selected as the power transfer amount for load redistribution.

8. The method for coordinated equalization of multi-motor load association in heavy-load UAV ESCs according to claim 7, characterized in that, The steps for generating the basic correction coefficients include: For load offset motors, a proportional allocation method based on the difference ratio is adopted. According to the proportion of the difference between the comprehensive load factor and the average load of each load offset motor in the total load excess, the drive duty cycle correction coefficient corresponding to each load offset motor is generated. For the collaborative compensation motors, a proportional allocation method based on the difference ratio is adopted. According to the proportion of the difference between the average load and the comprehensive load factor of each collaborative compensation motor in the total load margin, the drive duty cycle compensation coefficient corresponding to each collaborative compensation motor is generated. The generated correction and compensation coefficients are normalized based on the power transfer amount to ensure that the power reduction of the load offset motor is equal to the power increase of the cooperative compensation motor, thus satisfying the constraint that the overall output of the multi-motor system remains consistent.

9. The multi-motor load association method for heavy-load UAV ESC cooperative equalization according to claim 8, characterized in that, The dynamic adjustment and coefficient correction steps include: For load-off motors, the drive duty cycle correction coefficient is adjusted in segments according to the load change trend of each motor. Different adjustment ranges are used when the load change trend is within the preset value range. The greater the load change trend, the greater the adjustment range. For the collaborative compensation motor, the drive duty cycle compensation coefficient is adjusted differently based on the load type label obtained from the load trend analysis and type discrimination steps. When the load type label is attitude-coupled transient load, the compensation strength of the compensation coefficient is reduced, and when the load type label is continuous load, the compensation strength of the compensation coefficient is increased. The output is the final correction coefficient of the load offset motor and the final compensation coefficient of the collaborative compensation motor after adjustment.

10. The multi-motor load association method for heavy-load UAV ESC cooperative equalization according to claim 1, characterized in that, S4 includes: The final correction coefficient of the load offset motor and the final compensation coefficient of the cooperative compensation motor are mapped to the drive duty cycle of the corresponding ESC, respectively, to generate the drive duty cycle adjustment amount of each ESC. An exponential smooth transition function is used to continuously and smoothly adjust the drive duty cycle of each ESC to generate a smooth transition drive duty cycle. The smoothly transitioned drive duty cycle is output as a control signal to each corresponding ESC.

Images