An unmanned aerial vehicle control system and method based on a single-axis load cell

Abstract

This application provides a drone control system and method based on a single-axis load cell, comprising: an external force input handle at the first end of the drone's fuselage; a first single-axis load cell at the connection between the external force input handle and the fuselage; a second and a third single-axis load cell at the second end of the drone opposite to the first end, the second and third single-axis load cells being used to detect external forces in a second direction of the drone; a fourth single-axis load cell at the third end of the drone, the fourth single-axis load cell being used to detect external forces in a third direction of the drone; and a processing module for calculating the comprehensive external forces of the drone based on the detected values, and performing real-time control of the drone based on the comprehensive external forces. Low-cost, real-time stable control of the drone is achieved based on a lightweight, small-sized, and low-cost single-axis sensor.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) flight control technology, and in particular to a UAV control system and method based on a single-axis load cell. Background Technology

[0002] In recent years, the booming development of the low-altitude economy has fully demonstrated the enormous potential of drones in various application scenarios. Especially in collaborative operations with humans, such as drone cargo loading and unloading, drones, as aerial power platforms, can assist humans in completing complex production tasks such as load handling and equipment installation, significantly improving operational efficiency and flexibility. In such human-machine collaborative scenarios, quadcopter drones need to reach a designated location and hover during mission execution. Then, cargo is transferred from the cargo output location to the drone's transport platform. During cargo transfer, the drone is often subjected to external forces and torques exerted by operators or the external environment. These external forces are often difficult to describe using precise mathematical models, yet they significantly affect the dynamic characteristics of the drone, potentially leading to decreased flight stability or mission failure. Therefore, how to accurately estimate these unknown external forces and torques in real time, and thus, without human intervention, perform real-time adaptive compensation and response to the drone control system based on the estimation results, has become a key technical challenge to ensure the safe and efficient completion of collaborative tasks by quadcopter aircraft.

[0003] External force sensing is a fundamental requirement for adaptive measurement. Currently, among various technical approaches to achieve external force sensing, direct measurement methods have received widespread attention due to their high response speed and measurement accuracy. However, for small-sized, low-payload quadrotor UAV platforms, conventional multi-axis force sensors are not only structurally complex and heavy, but also relatively expensive to manufacture, making them difficult to integrate directly into such weight- and space-sensitive flight platforms. This increases the hardware and algorithm costs of UAV attitude stabilization control. Furthermore, while some existing indirect estimation methods based on dynamic models or filtering algorithms do not require external sensors, their estimation accuracy and robustness often fail to meet the requirements of high-reliability collaborative tasks when faced with unmodeled dynamics, parameter uncertainties, and complex external disturbances.

[0004] Therefore, there is an urgent need to develop a system for small quadcopter drones that can adaptively, quickly, and accurately sense external forces / torques in order to control the drones. Summary of the Invention

[0005] In view of this, this application provides a UAV control system and method based on a single-axis load cell to achieve the effect of low cost, low weight, and high accuracy of UAV external force measurement.

[0006] The first aspect of this application provides a UAV control system based on a single-axis load cell, characterized in that the system includes a UAV and multiple single-axis load cells.

[0007] The first end of the fuselage of the drone is provided with an external force input handle, which is used to allow external force to be input into the drone from the outside.

[0008] A first single-axis weighing sensor is provided at the connection between the external force input handle and the fuselage. The first single-axis weighing sensor is used to detect the external force in the first direction of the UAV.

[0009] The second end of the drone, opposite to the first end, is provided with a second single-axis load cell and a third single-axis load cell. The second single-axis load cell and the third single-axis load cell are used to detect the external force of the drone in a second direction, where the first direction is perpendicular to the second direction.

[0010] The third end of the drone is equipped with a fourth single-axis load cell, which is used to detect the external force in a third direction of the drone. The third direction is perpendicular to the first direction and the second direction.

[0011] The processing module is used to calculate the comprehensive external force of the UAV based on the detection values ​​of the first single-axis load cell, the second single-axis load cell, the third single-axis load cell, and the fourth single-axis load cell, and to perform real-time control of the UAV based on the comprehensive external force.

[0012] A second aspect of this application provides a UAV control method based on a single-axis load cell, the method being applied to the aforementioned system, the method comprising:

[0013] Acquire the detection values ​​of the first single-axis load cell, the second single-axis load cell, the third single-axis load cell, and the fourth single-axis load cell;

[0014] The external force information of the UAV is calculated by integrating the detected values; the control commands of the UAV are generated based on the external force information.

[0015] This application provides a UAV control system and method based on a single-axis load cell. This system enables direct measurement of the three-dimensional forces acting on the UAV when external forces are applied to a fixed external force input platform. It replaces traditional multi-axis force sensors with multiple single-axis sensors, achieving rapid and accurate force measurement while reducing the overall system complexity, manufacturing cost, and structural weight. Furthermore, the forces in multiple directions are numerically calculated to obtain the combined force information in each direction, directly addressing the force requirements for attitude stabilization control of the UAV. This eliminates the need for complex calculations of the components in each direction during real-time attitude control, improving the accuracy of attitude control, simplifying the computational load of UAV control, and significantly reducing the system's weight, size, and cost while maintaining measurement functionality. Attached Figure Description

[0016] Figure 1 A schematic diagram of the structure of an unmanned aerial vehicle (UAV) control system based on a single-axis weighing sensor provided in this application;

[0017] Figure 2 A flowchart of a UAV control method based on a single-axis weighing sensor provided in this application. Detailed Implementation

[0018] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0019] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0020] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0021] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0022] Figure 1 This is a schematic diagram of a UAV control system based on a single-axis load cell, provided in this application. Please refer to... Figure 1 The system provided in this embodiment includes a drone and multiple single-axis load cells.

[0023] The first end of the fuselage of the drone is provided with an external force input handle, which is used to allow external force to be input into the drone from the outside.

[0024] A first single-axis weighing sensor is provided at the connection between the external force input handle and the fuselage. The first single-axis weighing sensor is used to detect the external force in the first direction of the UAV.

[0025] The second end of the drone, opposite to the first end, is provided with a second single-axis load cell and a third single-axis load cell. The second single-axis load cell and the third single-axis load cell are used to detect the external force of the drone in a second direction, where the first direction is perpendicular to the second direction.

[0026] The third end of the drone is equipped with a fourth single-axis load cell, which is used to detect the external force in a third direction of the drone. The third direction is perpendicular to the first direction and the second direction.

[0027] The processing module is used to calculate the comprehensive external force of the UAV based on the detection values ​​of the first single-axis load cell, the second single-axis load cell, the third single-axis load cell, and the fourth single-axis load cell, and to perform real-time control of the UAV based on the comprehensive external force.

[0028] Combination Figure 1 As shown, the drone provided by this invention is typically a small quadcopter drone, which features a simple structure, light weight, and agile movement, and is often used in drone operation scenarios such as short-distance cargo transportation. Figure 1 The diagram shown is a design drawing of the drone's fuselage frame, which does not include the drone's rotor and shape, but only the main frame of the fuselage. Figure 1The far left is the drone's head position, and the external force input handle is installed at the drone's tail. The top of the external force input handle is used to place the transported goods during transport. The external force input handle is used to make contact between the goods placed on top of the handle and the drone, at which point the drone is in a hovering posture. Once the goods make contact with the external force input handle, the process is divided into the following stages based on the positional relationship between the goods and the handle: When contact is made, the majority of the goods' weight is still concentrated at the transport end, and the forces acting on the drone change instantaneously and significantly; subsequently, as the contact area between the goods and the handle increases, the mass of the drone increases, and the external force increases proportionally with the transport speed; when the contact area between the goods and the handle no longer changes, the goods are now entirely above the handle, specifically at the edge of the platform above the handle, i.e., a safe distance threshold from the boundary of the platform above the handle. During the process of transporting goods to the drone, the external forces acting on the handle are all centered on a single point on the boundary, with the magnitude and direction of the forces changing in real time. Existing force analysis technologies assess the external forces acting on a drone from its center of mass. No effective solution has been proposed for force analysis at a point on the edge. It can only be achieved under various assumptions through complex sensor detection and algorithmic calculation. The measurement accuracy is low, and the cost of the sensors used is high and the algorithmic calculation process is complex.

[0029] The drones mentioned in this invention are typically micro-drones, usually less than 15 centimeters in size and weighing less than 300 grams, possessing autonomous or semi-autonomous flight capabilities. Of course, drones can be of various sizes, and this invention is not limited to these; further details will not be elaborated upon here. Microelectromechanical systems (MEMS) technology can be used on the drones to integrate sensors, processors, and power modules, achieving lightweighting and miniaturization. The flight control unit integrates sensors such as gyroscopes, accelerometers, barometers, cameras, and infrared sensors. It can also be equipped with a small robotic arm to perform attitude calculation, obstacle avoidance, path planning, and operational control. Different sensors and actuators can be combined according to mission requirements to adapt to different task needs.

[0030] Combination Figure 1As can be seen, the first end of the drone's fuselage is equipped with an external force input handle, specifically the tail end. The external force input platform is T-shaped. The external force input handle is used to input external force to the drone. The T-shaped external force input handle includes a horizontal part and a vertical part. The horizontal part is used to receive external force input and place goods, while the vertical part is used to connect the drone and the horizontal part. In this shape, once the goods start to contact the external force input handle, the force-bearing end is the long side of the horizontal part away from the drone's fuselage, that is, the force-bearing point is the farthest end of the external force input platform in the direction away from the drone's fuselage.

[0031] The UAV has four single-axis load cells (ABCD) on its fuselage. Each sensor measures the change in force in only one direction, called the sensitive axis direction. It is a device specifically designed to detect and measure tensile or compressive forces applied along a single, specific axis and convert them into quantifiable electrical signals. The four single-axis load cells are orthogonally distributed. Each load cell includes a top platform facing the sensitive axis direction. The operation of the single-axis load cell relies on the elastic deformation effect of the metal and the resistance change effect of the strain gauge. It includes at least an elastic body, a strain gauge, and a signal conversion and processing module. The strain gauge is connected to the detection circuit of the signal conversion and processing module. The detection circuit is typically a bridge circuit structure, such as a Wheatstone bridge. The strain gauge is attached to the deformation center region of the elastic body. The elastomer, typically made of high-strength alloy steel, aluminum, or stainless steel, undergoes a minute deformation proportional to the applied force when an external force is applied along the sensitive axis of the uniaxial load cell. The strain gauge deforms accordingly at its deformation center, causing a change in its resistance. A Wheatstone bridge detects this resistance change and converts it into a millivolt-level voltage signal proportional to the applied force. The detection circuit connects to a signal processing circuit, which includes at least an amplifier, an analog-to-digital converter, and a processor. The amplifier amplifies the millivolt-level voltage signal, which is then converted into a digital signal by the analog-to-digital converter. Finally, the processing module calculates and interprets the signal to obtain a precise force value.

[0032] A first single-axis load cell D is provided at the connection between the external force input handle and the fuselage. This first single-axis load cell detects the external force in a first direction of the drone. A second single-axis load cell B and a third single-axis load cell C are provided at the second end of the drone opposite to the first end. These two sensors detect the external force in a second direction, perpendicular to both the first and second directions. A fourth single-axis load cell is provided at the third end of the drone. This fourth sensor detects the external force in a third direction, perpendicular to both the first and second directions. Specifically, a spatial coordinate system is established with the center of mass of the drone as the origin, encompassing the drone's fuselage and the external force input platform. The plane containing the drone's fuselage is the xy-plane, and the vertically upward direction is the z-axis. The first direction is the y-axis, the second direction is the x-axis, and the third direction is the z-axis. The first single-axis load cell detects the external force in the y-axis direction, with the top platform facing the y-axis. The second and third single-axis load cells detect the external force in the x-axis direction, with the top platform facing the x-axis. The fourth single-axis load cell detects the external force in the z-axis direction, with the top platform facing the z-axis.

[0033] From their appearance, the four single-axis load cells are completely identical, such as... Figure 1As shown, the structure is specifically cylindrical, with the top platform being the top surface of the cylinder. The centroids of the first, second, and third single-axis load cells are in the same plane. The second and third single-axis load cells are coaxially arranged and symmetrical along a line passing through the centroid of the first single-axis load cell and parallel to the y-axis. Preferably, the centroids of the second and third single-axis load cells are located in the xy-plane, or they can be planes parallel to the xy-plane. Preferably, the centroid of the first single-axis load cell can be located on the y-axis. In this case, the centroids of the second and third single-axis load cells are symmetrical along the y-axis, and the second and third single-axis load cells are arranged facing each other, i.e., their top platforms are opposite each other. Preferably, the second single-axis load cell B is used to detect external forces in the x-axis direction, and the third single-axis load cell C is used to detect external forces in the opposite x-axis direction. Sensor D is used to detect external forces in the y-axis direction, and the fourth single-axis load cell A is used to detect external forces in the positive z-axis direction. The center of mass of the fourth single-axis load cell A is coaxially mounted with the center of mass of the UAV. Therefore, the x-axis coordinate of the first single-axis load cell is 0, the y-axis coordinate is negative, and the z-axis coordinate is 0; the x-axis coordinate of the second single-axis load cell is negative, the y-axis coordinate is positive (the absolute value is defined as a fixed design value L), and the z-axis coordinate is 0; the x-axis coordinate of the third single-axis load cell is positive, the y-axis coordinate is positive, and the z-axis coordinate is 0; the x-axis coordinate of the fourth single-axis load cell is 0, the y-axis coordinate is 0, and the z-axis coordinate is negative. The second and third single-axis load cells are located in the xy plane of a three-dimensional coordinate system established with the UAV fuselage as a reference, and the second and third single-axis load cells are symmetrical along the y-axis. The first single-axis sensor is equidistant from the second and third single-axis sensors, and is greater than the first single-axis sensor from the fourth single-axis sensor. In other words, the absolute value of the y-axis coordinate of the first single-axis load cell is less than the absolute value of the z-axis coordinate of the fourth single-axis load cell.

[0034] like Figure 1As shown, the UAV provided in this embodiment of the invention adopts a quadcopter layout with a symmetrical cross-shaped structure at its center. The central fuselage is a cuboid structure, serving as the main load-bearing structure of the system and integrating core components such as the flight controller, battery, and processing module. The first end is the tail end of the central fuselage, with a first recessed portion. The first recessed portion matches the size of the first single-axis load cell, and the external force input platform is fixedly connected to the first recessed portion, connecting to the fuselage of the UAV through the first recessed portion. The second end is the position of the central fuselage opposite to the first end, specifically the head of the central fuselage. The head is cuboid in shape, perpendicular to the central fuselage, and oriented at a 90° angle to the central fuselage. The central fuselage is positioned along the y-axis, and the head is positioned along the x-axis. Specifically, from the front view, the front of the central fuselage with its long side faces directly forward, and the side with its short side faces to the sides. The side of the head with its short side also faces directly forward. The second end of the drone has a second recess and a third recess arranged opposite each other along the x-axis direction, that is, the second and third recesses are provided at the two side ends of the head. The second single-axis load cell is installed in the second recess, and the third single-axis load cell is installed in the third recess. The first and second recesses are respectively sized to match the first and second single-axis load cells. Preferably, the dimensions of the first, second, and third recesses are exactly the same, and the dimensions of the first, second, and third single-axis load cells are also exactly the same. The fourth single-axis load cell is installed on the bottom plane of the drone's fuselage. The four single-axis load cells provided by this invention utilize the geometric characteristics of a cross-shaped structure to calculate three-dimensional external forces without the need for complex calculation algorithms. Accurate three-directional external forces can be obtained through a single measurement and signal processing, eliminating the need for assumptions about the flight scenario and force decomposition. The external force component applied via the handle along the y-axis of the drone is directly detected, while the external force along the x-axis is measured jointly by two opposing single-axis load cells, improving measurement stability and accuracy. The processor calculates the combined three-dimensional external force acting on the drone based on the individual detection values. This combined external force is represented by force vectors in the three directions, facilitating real-time stable attitude control of the drone's suspended fuselage based on the combined external force. The method provided by this invention places the four single-axis load cells inside the drone fuselage through a recessed design, resulting in no additional components from the outside and saving space occupied by the drone's fuselage.

[0035] After quickly obtaining real-time external force results, the method provided by this invention eliminates the need to fuse the forces of various objects. Using the real-time force situation of the drone itself as a basis, the processing module directly uses the forces in each direction without conversion or disassembly, enabling rapid control of the drone. Specifically, the drone also includes a processing module used to calculate the comprehensive external force of the drone based on the detection values ​​of the first, second, third, and fourth single-axis load cells. The processing module includes at least an analog-to-digital conversion circuit and a processing circuit.

[0036] As an optional embodiment, there can be one or more analog-to-digital conversion circuits. If there is only one analog-to-digital conversion circuit, the sensor values ​​are sequentially input into the circuit to convert the continuous analog voltage signal output by the single-axis load cell into a discrete digital code, enabling it to be recognized and processed by the microprocessor. If there are multiple analog-to-digital conversion circuits, each corresponding to one single-axis load cell, each load cell detects an analog pressure signal and inputs it to its corresponding circuit. Each analog pressure signal is then converted into a discrete numerical signal in parallel. The principle of the analog-to-digital conversion circuit is a common signal processing circuit in the prior art, and its structure and processing flow will not be elaborated upon here.

[0037] The detection values ​​from all single-axis load cells are ultimately transmitted to the processing module. The processing module filters the detection values ​​from the four single-axis load cells and calculates the combined external force acting on the UAV. Based on this combined external force, the UAV is controlled in real time. As an optional embodiment, the processing module at least includes a filtering module. Filtering is performed before calculating the final combined external force. The filtering module has filtering functionality, and optionally, the filtering parameter can be median filtering. The filtering module determines the filtering parameters based on the order in which the detection values ​​from the four single-axis load cells arrive at the filtering module. The filtering parameters differ for different single-axis load cells. The filtering module determines the median filtering parameters sequentially based on the order of the detection values ​​from the single-axis load cells, and filters the detection values ​​according to these parameters. The filtering parameters for each single-axis load cell are determined before the pressure detection signal arrives. The determination of the filtering parameters will be elaborated below and will not be repeated here. In the UAV external force measurement system, the detection signals from each single-axis load cell may introduce impulse noise and random interference. To improve the quality of the detected pressure signal, this invention uses median filtering to preprocess the pressure signal sequence. Median filtering is applied to the continuous sampling sequence output by each sensor to remove transient pulse interference. Let the sensor sampling sequence be S=[s1,s2,...,s...]. n If the filter window size is m (an odd number), then the median filter output is: .in, This is the filtered value at the i-th sampling point. For the detected analog pressure signal, the original pressure signal of each single-axis weighing sensor is preprocessed in real time, effectively suppressing the instantaneous interference pulses experienced by the UAV in complex electromagnetic and vibration environments. This provides a foundation for the subsequent accurate calculation of the three-dimensional comprehensive external force, improving the accuracy and robustness of the entire measurement system.

[0038] As an optional embodiment, the weighing sensor is a miniature weighing sensor, and the analog-to-digital conversion module is a small-sized analog-to-digital conversion module based on the CS1237 chip. The combined external force includes at least a combined external force value and a combined external force torque. The combined external force value includes force values ​​in the x, y, and z directions; similarly, the combined external force torque also includes torques in the x, y, and z directions. Specifically, the filtered value is [ ,in, The measurements are from the x-axis sensors, specifically the second and third single-axis load cells. The measurement result is from the y-axis sensor, i.e., the first single-axis load cell. This is the measurement result from the z-axis sensor, i.e., the fourth single-axis load cell. At this point, the combined external force is specifically the x-axis combined external force. ;Comprehensive external force along the y-axis Z-axis combined external force The resultant external force along the x-axis is the sum of the external forces detected by the second and third single-axis load cells. If the two forces are in opposite directions, the sum is the difference between the two forces, and the direction is the direction of the force with the larger absolute value. If the two forces are in the same direction, the sum is the sum of the two forces, and the direction is the direction in which the two forces are located. The y and z directions are detected by a single single-axis load cell. Therefore, the combined external force in these two directions is equal to the detection value of the single-axis load cell.

[0039] Furthermore, the combined external torque in the three directions is as follows:

[0040] x-axis combined external force torque ;

[0041] Combined external force torque along the y-axis ;

[0042] z-axis combined external force torque ;

[0043] in, The total external force value along the x-axis. This represents the combined external force value along the y-axis. This represents the combined external force value along the z-axis; , Let x and y be the coordinates of the fourth single-axis load cell relative to the center of mass of the UAV. , The y-axis and z-axis coordinates of the second or third single-axis load cell relative to the center of mass of the UAV. , Let x and z be the coordinates of the first single-axis load cell relative to the center of mass of the UAV.

[0044] As a preferred embodiment, if, as described above, the first single-axis load cell has an x-axis coordinate of 0, a negative y-axis coordinate, and a z-axis coordinate of 0; the second single-axis load cell has a negative x-axis coordinate, a positive y-axis coordinate with an absolute value defined as a fixed design value L, and a z-axis coordinate of 0; the third single-axis load cell has a positive x-axis coordinate, a positive y-axis coordinate, and a z-axis coordinate of 0; and the fourth single-axis load cell has a x-axis coordinate of 0, a y-axis coordinate of 0, and a negative z-axis coordinate. In this case, the calculation of the aforementioned combined external force torque becomes:

[0045] x-axis combined external force torque ;

[0046] Combined external force torque along the y-axis ;

[0047] z-axis combined external force torque Where L is the distance in the y direction from the x-axis sensor to the z-axis.

[0048] After obtaining force and torque information in three directions, the processing module is further used to perform real-time control of the UAV based on the combined external forces. The module determines corresponding control parameters based on the real-time combined external forces and torques in the three directions. As an optional embodiment, the control parameter is the change in the influence of external forces in any direction, such as flight altitude or horizontal turning angle. For each control parameter, a corresponding compensation control command is calculated based on the combined external forces and torques in the corresponding direction to achieve real-time control of the UAV and maintain its stable attitude. The compensation control command is used to compensate for the influence margin of the combined external forces and torques on the UAV's attitude.

[0049] The UAV control device based on single-axis load cells provided by this invention achieves a comprehensive effect of high-accuracy measurement of external forces in three-dimensional space for UAVs under the premise of low cost and low weight. It effectively solves the problems of complexity, high cost, and heavy weight of traditional multi-axis force sensor systems, as well as the low measurement accuracy and complex algorithms of existing force analysis methods in edge point contact scenarios. Furthermore, it can control the attitude of UAVs in a low-cost and low-weight manner, maintaining the attitude stability of UAVs without human intervention. Specifically, multiple orthogonally distributed single-axis load cells are used to replace multi-axis force sensors, which detect external forces in three vertical directions respectively, constructing a low-cost and low-weight three-dimensional force measurement structure. The single-axis sensor itself has a simple structure and low cost. Its orthogonal layout directly realizes the measurement of three-dimensional external forces in different directions, avoiding the use of expensive and bulky multi-axis sensors, and significantly reducing the overall complexity, manufacturing cost and structural weight of the system. In terms of positioning, four single-axis load cells are orthogonally distributed in a specific geometric relationship and integrated into a recess in the UAV fuselage. This not only allows for direct position selection but also... The method achieves efficient calculation of spatial forces, simplifies the numerical calculation process, and embeds the sensors within the fuselage, saving external space and maintaining the miniaturization and lightweight characteristics of the UAV. At the same time, the processing module can quickly obtain the comprehensive external force value and comprehensive external torque of the UAV in three directions by directly summing the detection values ​​of the four single-axis sensors, eliminating the complex force decomposition and algorithm calculation process in the prior art. It directly converts the measurement results into force information in various directions required for UAV attitude control, thereby simplifying the calculation of force analysis and improving the response speed and accuracy of attitude control.

[0050] Figure 2 A flowchart of a UAV control method based on a single-axis weighing sensor provided in this application is shown. The method is applied to the system described in Embodiment 1, and the method includes:

[0051] S101. Obtain the detection values ​​of the first single-axis load cell, the second single-axis load cell, the third single-axis load cell, and the fourth single-axis load cell. The first to fourth single-axis load cells are all unidirectional load cells, and their detection values ​​are the external pressure values ​​applied in a single direction.

[0052] Specifically, before calculating the external force information of the UAV by fusing the detected values, the method further includes: receiving the measurement values ​​of each single-axis load cell, specifically, each single-axis load cell includes two x-axis load cells and y and z-axis load cells, and the measurement values ​​of each single-axis load cell are the external forces on the UAV in each direction measured at the same time; determining the filter value of each single-axis load cell based on the distance of each single-axis load cell from the external force input handle and the flight state of the UAV, specifically, the filtering method is median filtering, and the size of the filtering window is determined based on the distance and flight state as the filter value;

[0053] The measured value is filtered according to the filter value of each single-axis weighing sensor to obtain the correction value; the correction value is then converted from analog to digital to obtain the detection value.The system determines the filter value in real time based on the UAV's flight state, providing an adaptive filtering mechanism that allows the system to intelligently adjust the filter parameters according to dynamic changes in the flight state, thereby achieving an optimal balance between measurement accuracy and response speed. Preferably, the system determines the UAV's real-time flight state and, based on the real-time flight state, determines the environmental interference influence factor of the UAV's fuselage. The system continuously monitors real-time data from the flight control unit, including but not limited to acceleration, angular velocity, rotor speed, and other state data. Based on the state data, the system identifies the UAV's real-time flight state, which includes at least the climb phase, stable flight phase, and landing phase. Based on the real-time flight state, the system determines the direction of environmental interference forces and, based on the environmental interference... The direction of environmental interference forces affects the target single-axis load cell. For example, during the landing phase, based on its flight path, the main directions of environmental interference forces it experiences are the z-axis and x-axis. At this time, the single-axis load cells respond in these two directions. There can be one or four target single-axis load cells. The airframe environmental interference influence factor of the target single-axis load cell is determined based on the real-time flight status. The flight parameters of the UAV in that direction are obtained based on the detection force direction of the target single-axis load cell. These flight parameters are input into the UAV simulation model to predict the interference force value in that direction. The interference force value is normalized and used as the airframe environmental interference influence factor. This factor indicates the degree of environmental influence on the target single-axis load cell. A larger factor indicates greater environmental interference with the sensor's detection. In this case, the filter window size should be appropriately increased to relax the filtering strategy, avoiding signal distortion caused by single interference and ensuring the stability of the force signal. Based on the distance of each single-axis load cell relative to the external force input handle, the adjustment margin of the environmental interference influence factor is determined, resulting in a filter adjustment factor for each single-axis load cell. The adjustment margin represents the degree of environmental interference influence due to the force propagation distance, and is a normalized value of the distance. Specifically, it is the ratio of the distance of the current single-axis load cell from the external force input handle to the distance of all single-axis load cells from the external force input handle. The ratio of the maximum distance of the input handle, the product of the environmental interference impression factor and the adjustment margin of each single-axis load cell, is used as the filter adjustment factor. This allows for adjustment of the filter value based on distance and flight status, ensuring that the filter parameters match the environment and force propagation interference when external forces are applied. Based on the product of the filter adjustment factor and the basic filter value, the filter value of each single-axis load cell is calculated. Specifically, the product of the adjustment factor and the basic filter value is used as the adjustment scale for the filter window size. With the basic filter value as the center and the adjustment scale as the step size, the filter window size is adjusted. The difference between the adjusted filter window size and the window size of the basic filter value is used as the adjustment scale. The adjusted filter window is centered on the center of the original filter window.Specifically, the base filter value is the window size for median filtering. This window size can be a preset value, the filter value at the previous filtering time, or the filter value for a drone in the same flight state. Since the installation positions of the individual axis load cells on the fuselage differ, under the same environment, the force needs to travel a certain path for propagation. Therefore, the influence of vibration and force patterns at the detection points of different sensors also differs. Specifically, the closer to the external force input handle, the shorter the force propagation path and the less the influence of the environment during propagation, thus resulting in a smaller adjustment margin. The method provided by this invention, through a dynamically adaptive filtering strategy, ensures that a smooth, stable, and rapidly reflective detection signal reflecting real external force changes can be obtained in all flight scenarios, laying a reliable data foundation for the subsequent accurate calculation of the three-dimensional comprehensive external force.

[0054] As an optional implementation, after applying median filtering, the median filter can be further corrected by calculating the corrected value. To further improve estimation accuracy, reduce noise, and predict external force signals, a Kalman filter is applied to the corrected value sequence of each single-axis weighing sensor channel to perform independent single-axis external force state estimation. Each single-axis external force state is defined as follows: For the k-th single-axis sensor (k=1,2,3,4, corresponding to the x, y, and z directions respectively), the state vector is defined as... Among them, F k F is the current calculated real-time estimate of the external force along the sensor axis. dot_k This represents the first derivative (rate of change) of the real-time external force. The state is updated based on the Kalman filter recursive equation. First, the state prediction is calculated based on the following formula: Where A is the state transition matrix, A = [[1, d], t ], [0,1]],d t This is the filtering period. This is the posterior state estimate from the previous time step. First, estimate the prior state at the current moment. Second, calculate the covariance based on the following formula: Where P is the state estimation error covariance matrix and Q is the process noise covariance matrix, used to model the randomness of external force variations. Finally, the Kalman gain is calculated based on the following formula: .

[0055] Where H is the observation matrix, H = [1, 0], indicating that only the external force value F is observed. k R k(t) For adaptive observation noise covariance. Finally, the state update is (measurement update):

[0056] ;

[0057] Among them, z k(t) This is the detection value obtained from the correction value after analog-to-digital conversion at the current moment.

[0058] Covariance update:

[0059] ;

[0060] After the above steps, the final output F k (Right now This refers to the more accurate and smoother detection value in the single-axis direction after Kalman filtering optimization, which is used for subsequent external force information fusion calculation.

[0061] The R matrix here changes according to the drone's state. Its specific value is R = α * v 2 Where v is the flight speed of the UAV along this axis; α is a constant coefficient. The greater the flight speed of the UAV along this axis, the lower the reliability of its measurement value, and the larger the coefficient R. Conversely, the smaller the speed of the UAV along this axis, the higher the reliability of the measurement, the smaller the coefficient R, and the greater the impact on the update of the estimated value.

[0062] S102. Calculate the external force information of the UAV by fusing the detected values; generate the UAV control command based on the external force information.

[0063] The sum of the detection values ​​of single-axis load cells in the same direction is used to calculate the resultant external force values ​​in each direction; the coordinates of each single-axis load cell in a three-dimensional coordinate system established with the UAV fuselage as the reference are obtained; the distance of the coordinates from the origin of the three-dimensional coordinate system in the same direction as each detection value is calculated; and the resultant external torque corresponding to each resultant force is calculated based on the product of the distance and the corresponding force. Specifically, the filtered values ​​are [ ,in, The measurements are from the x-axis sensors, specifically the second and third single-axis load cells. The measurement result is from the y-axis sensor, i.e., the first single-axis load cell. This is the measurement result from the z-axis sensor, i.e., the fourth single-axis load cell. At this point, the combined external force is specifically the x-axis combined external force. ;Comprehensive external force along the y-axis Z-axis combined external force The resultant external force along the x-axis is the sum of the external forces detected by the second and third single-axis load cells. If the two forces are in opposite directions, the sum is the difference between the two forces, and the direction is the direction of the force with the larger absolute value. If the two forces are in the same direction, the sum is the sum of the two forces, and the direction is the direction in which the two forces are located. The y and z directions are detected by a single single-axis load cell. Therefore, the combined external force in these two directions is equal to the detection value of the single-axis load cell.

[0064] Furthermore, the combined external torque in the three directions is as follows:

[0065] x-axis combined external force torque ;

[0066] Combined external force torque along the y-axis ;

[0067] z-axis combined external force torque ;

[0068] in, The total external force value along the x-axis. This represents the combined external force value along the y-axis. This represents the combined external force value along the z-axis; , Let x and y be the coordinates of the fourth single-axis load cell relative to the center of mass of the UAV. , The y-axis and z-axis coordinates of the second or third single-axis load cell relative to the center of mass of the UAV. , Let x and z be the coordinates of the first single-axis load cell relative to the center of mass of the UAV.

[0069] As a preferred embodiment, if, as described above, the first single-axis load cell has an x-axis coordinate of 0, a negative y-axis coordinate, and a z-axis coordinate of 0; the second single-axis load cell has a negative x-axis coordinate, a positive y-axis coordinate with an absolute value defined as a fixed design value L, and a z-axis coordinate of 0; the third single-axis load cell has a positive x-axis coordinate, a positive y-axis coordinate, and a z-axis coordinate of 0; and the fourth single-axis load cell has a x-axis coordinate of 0, a y-axis coordinate of 0, and a negative z-axis coordinate. In this case, the calculation of the aforementioned combined external force torque becomes:

[0070] x-axis combined external force torque ;

[0071] Combined external force torque along the y-axis ;

[0072] z-axis combined external force torque ; where L is the y-direction distance from the x-axis sensor to the z-axis.

[0073] After calculating the external force information of the UAV by fusing the detected values, the method further includes: calculating the attitude change based on the external force information; calculating a correction command based on the attitude change; controlling the attitude of the UAV according to the correction command, so that the real-time attitude of the UAV after control meets the balance requirements.

[0074] The three-dimensional net external force acting on the center of mass of the UAV is obtained through the aforementioned steps. With three-dimensional resultant external torque These two factors together constitute complete, real-time updated external force information. Based on this, adaptive control commands can be generated, the core idea of ​​which is to use this external force information for feedforward compensation.

[0075] The control command is derived from the feedback control command U. fb and feedforward compensation command U ff Superimposed generation. Feedback control command U fb The feedforward compensation command U is calculated from state errors such as position, attitude, velocity, and angular velocity using a PID controller and is used for basic stability and trajectory tracking. ff Then directly from the measured external force information (F) ext M ext The calculations are intended to actively counteract disturbances. For torque disturbances, a reverse feedforward torque command is directly generated:

[0076] ;

[0077] Superimposed on roll, pitch, and yaw channels, where K M This is an adjustable diagonal gain matrix. For force disturbances, compensation needs to be mapped to attitude and lift commands. The horizontal disturbance force is calculated as follows:

[0078] ;

[0079] Compensation is achieved by calculating the desired feedforward tilt angle:

[0080] ;

[0081] This angle, superimposed on the original desired attitude angle, serves as the input to the attitude controller, thereby guiding the UAV to actively tilt to generate counterforce; vertical disturbance force F z Then, compensation is directly achieved through feedforward lift commands:

[0082] U1 ff = -K z * F z ;

[0083] Gain K M and Kz It can adaptively adjust based on flight status (such as speed and altitude) or disturbance characteristics (such as frequency and amplitude) to optimize dynamic response in different scenarios. Ultimately, the synthesized command for all channels is...

[0084] U total = U fb + U ff ;

[0085] This method enables UAVs to quickly and actively counteract external disturbances through force measurement feedforward compensation, significantly improving flight stability and control accuracy in complex environments (such as gusts, load changes, and external force manipulation).

[0086] The method in this embodiment can be used... Figure 1 The specific implementation principle and process of the system shown in the embodiment are similar, and will not be repeated here.

[0087] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A drone control system based on a single axis load cell, characterized in that, The system includes a drone and multiple single-axis load cells. The first end of the fuselage of the drone is provided with an external force input handle, which is used to allow external force to be input into the drone from the outside. A first single-axis weighing sensor is provided at the connection between the external force input handle and the fuselage. The first single-axis weighing sensor is used to detect the external force in the first direction of the UAV. The second end of the drone, opposite to the first end, is provided with a second single-axis load cell and a third single-axis load cell. The second single-axis load cell and the third single-axis load cell are used to detect the external force of the drone in a second direction, where the first direction is perpendicular to the second direction. The third end of the drone is equipped with a fourth single-axis load cell, which is used to detect the external force in a third direction of the drone. The third direction is perpendicular to the first direction and the second direction. The processing module is used to calculate the comprehensive external force of the UAV based on the detection values ​​of the first single-axis load cell, the second single-axis load cell, the third single-axis load cell, and the fourth single-axis load cell, and to perform real-time control of the UAV based on the comprehensive external force.

2. The system according to claim 1, characterized in that, A three-dimensional coordinate system is established with the center of mass of the UAV as the origin and the fuselage plane of the UAV as the xy-axis plane. The first direction is the y-axis, the second direction is the x-axis, and the third direction is the z-axis.

3. The system according to claim 1, characterized in that, The second single-axis load cell and the third single-axis load cell are located in the xy plane of a three-dimensional coordinate system established with the UAV fuselage as a reference, and the second single-axis load cell and the third single-axis load cell are symmetrical along the y-axis.

4. The system according to claim 1, characterized in that, The external force input handle is fixedly connected to the machine body through a first recess, and the first single-axis load cell is installed in the first recess. The size of the first recess matches that of the first single-axis load cell. The second end of the UAV is provided with a second recess and a third recess along the x-axis direction. The second single-axis load cell is installed in the second recess, and the third single-axis load cell is installed in the third recess. The first recess and the second recess are respectively matched with the size of the first single-axis load cell and the second single-axis load cell. The fourth single-axis load cell is installed on the bottom plane of the drone's fuselage.

5. The system according to claim 1, characterized in that, The first single-axis sensor is equidistant from the second and third single-axis sensors, and is greater than the first single-axis sensor from the fourth single-axis sensor.

6. A UAV control method based on a single-axis load cell, the method being applied to the system as described in any one of claims 1-5, the method comprising: Acquire the detection values ​​of the first single-axis load cell, the second single-axis load cell, the third single-axis load cell, and the fourth single-axis load cell; The external force information of the UAV is calculated by integrating the detected values; the control commands of the UAV are generated based on the external force information.

7. The method of claim 6, wherein, The process of fusing the detected values ​​to calculate the external force information of the drone includes: The resultant external force values ​​in each direction are calculated based on the sum of the detection values ​​of a single-axis load cell in the same direction. Obtain the coordinates of each single-axis weighing sensor in a three-dimensional coordinate system established with the UAV fuselage as the reference; Based on the coordinates, calculate the distance between the coordinates and the origin of the three-dimensional coordinate system in the same direction as each detection value; The resultant external torque corresponding to each resultant force is calculated based on the product of the distance and the corresponding force.

8. The method of claim 6, wherein, Before calculating the external force information of the UAV by fusing the detected values, the method further includes: Receives the measurement values ​​from each single-axis load cell; The filter value of each single-axis load cell is determined based on the distance of each single-axis load cell from the external force input handle and the flight state of the UAV. The corresponding measured values ​​are filtered based on the filter values ​​of each single-axis load cell to obtain the correction values; The correction value is converted from analog to digital to obtain the detection value.

9. The method of claim 6, wherein, After calculating the external force information of the UAV by fusing the detected values, the method further includes: Calculate the attitude change based on the external force information; Calculate the correction command based on the attitude change; The attitude of the UAV is controlled according to the correction command, and the real-time attitude of the UAV after control meets the balance requirements.

10. The method of claim 8, wherein, The step of determining the filter value of each single-axis load cell based on the distance of each single-axis load cell from the external force input handle and the flight state of the UAV includes: Determine the real-time flight status of the drone; The environmental interference impact factor of the UAV is determined based on the real-time flight status. Based on the distance of each single-axis load cell relative to the external force input handle, the adjustment margin of the environmental interference influence factor of the machine body is determined, and the filter adjustment factor corresponding to each single-axis load cell is obtained. The filter value of each single-axis weighing sensor is calculated based on the product of the filter adjustment factor and the basic filter value.

Images