Robot reconfigurable method and system based on multi-sensor fusion

Abstract

The invention provides a robot reconfigurable method and system based on multi-sensor fusion, and belongs to the field of robot control, and the method comprises the steps: dividing a robot mechanical body into a plurality of reconfigurable motion execution units; using a random sampling method to calibrate any reconfigurable motion execution unit to obtain an initial model, and calibrating a connection structure between two adjacent reconfigurable motion execution units to obtain an optimization model; performing paving operation on the optimization model in a plurality of movable directions to obtain an equivalent mathematical model; acquiring real-time monitoring data of the angle sensor and the gyroscope, calculating actual motion parameters, and inputting the real-time monitoring data into the equivalent mathematical model to obtain a simulation value; and fault judgment is conducted on the robot mechanical body based on the actual motion parameters and the simulation values, when the robot mechanical body breaks down, the parameters of the equivalent mathematical model are updated, the motion track is planned again, and real-time monitoring and track planning are repeated till motion control over the robot mechanical body is completed.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a robot reconfiguration method and system based on multi-sensor fusion. Background Art

[0002] With the development of advanced concepts such as intelligent manufacturing, industrial Internet and smart cities, reconfigurable systems and digital twin technologies as key enabling technologies have received strong support from the government and enterprises. Among them, digital twin technology, as an important means to achieve intelligent manufacturing, has been widely concerned by academia and the business community for its practical application, while reconfigurable systems are one of the key technologies for achieving the goal of intelligent manufacturing. The development of these technologies is of great significance to promoting the intelligent transformation of the manufacturing industry.

[0003] Traditional intelligent control systems have obvious deficiencies in adaptive capabilities, making it difficult to flexibly adjust system configurations according to changes in the environment and task requirements. At the same time, they also face technical challenges in realizing external digital twins of the system. Existing systems lack effective control methods for the smallest units, making it difficult to achieve flexible reconstruction of the system. In addition, the control accuracy of digital twin technology is insufficient, which limits the flexibility and adaptability of the system.

[0004] Therefore, finding a method that can flexibly adjust the system configuration according to changes in the environment and task requirements, and accurately control the digital twin model, is a technical problem that needs to be urgently solved by technical personnel in this field. Summary of the invention

[0005] The present invention provides a robot reconfiguration method and system based on multi-sensor fusion, which is used to solve the defect in the prior art that it is difficult to flexibly adjust the system configuration according to changes in the environment and task requirements, and realize the reconfigurability and self-adaptation ability of the robot mechanical body.

[0006] The present invention provides a robot reconfiguration method based on multi-sensor fusion, comprising the following steps:

[0007] S1. Divide the robot mechanical body into a plurality of reconfigurable motion execution units, each of which includes a joint, a motor, an angle sensor disposed on the input side of the motor, and a gyroscope disposed on the output side of the motor, wherein the angle sensor is concentrically connected to the motor, and a measuring axis of the gyroscope is tangent to the output side of the motor;

[0008] S2. Use a random sampling method to calibrate the connection order and connection mode of any reconfigurable motion execution unit to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model;

[0009] S3, performing paving operations on the optimized model in multiple movable directions to obtain an equivalent mathematical model;

[0010] S4, obtaining real-time monitoring data of the angle sensor and the gyroscope and calculating the actual motion parameters, and inputting the real-time monitoring data into an equivalent mathematical model to obtain a simulation value;

[0011] S5. Perform fault judgment on the robot mechanical body based on the actual motion parameters and simulation values. When the robot mechanical body fails, update the parameters of the equivalent mathematical model and re-plan the motion trajectory, and repeat steps S4-S5 until the robot mechanical body completes motion control.

[0012] According to a robot reconfiguration method based on multi-sensor fusion provided by the present invention, the robot mechanical body is divided according to the drive, reducer, connection structure of each output part in the robot mechanical body, components of the connection structure, each end-effector in the robot mechanical body and part of the end-effector, and the two adjacent reconfigurable motion execution units are connected by a connection structure.

[0013] According to a robot reconfiguration method based on multi-sensor fusion provided by the present invention, step S2 specifically includes:

[0014] S21, collecting data of the angle sensor in each reconfigurable motion execution unit, constraining the spatial structural motion of the robot mechanical body to the motion of the reconfigurable motion execution unit, and confirming the priority of each reconfigurable motion execution unit in the robot mechanical body, using the least squares method to process the data of the angle sensor to obtain an initial model of the robot mechanical body;

[0015] S22, calibrating step by step from the end of the mechanism of the robot mechanical body toward the base direction until the reconfigurable motion execution units of all degrees of freedom in the robot mechanical body are calibrated to obtain an optimized model of the robot mechanical body;

[0016] S23, determining the initial angle range of two adjacent reconfigurable motion execution units and the length of the rod connecting the two adjacent reconfigurable motion execution units according to the structural constraints of the robot mechanical body;

[0017] S24, determining an optimization model according to the initial angle range and the length of a rod connecting two adjacent reconfigurable motion execution units.

[0018] According to a robot reconfiguration method based on multi-sensor fusion provided by the present invention, step S22 of performing step-by-step calibration from the end of the mechanism of the robot mechanical body toward the base direction includes:

[0019] Step 1, recording the position of the reconfigurable motion execution unit to be calibrated in the initial model, and recording it as the zero point position;

[0020] Step 2: Control the reconfigurable motion execution unit to be calibrated to perform reverse growth motion until the return data of the sensor in the reconfigurable motion execution unit is 0, and the reconfigurable motion execution unit to be calibrated reaches the actual motion maximum position, and the actual motion maximum position is recorded;

[0021] Step 3: Determine the midpoint between the actual maximum motion position and the zero point position, perform a binary search between the midpoint position and the mechanical limit position of the reconfigurable motion execution unit in the robot mechanical body, and obtain the exact position of the reconfigurable motion execution unit of the degree of freedom to be calibrated.

[0022] According to a robot reconfiguration method based on multi-sensor fusion provided by the present invention, the paving operation specifically includes:

[0023] Taking the base reconfigurable motion execution unit of the robot mechanical body as the starting point, setting the motion reference direction, driving the motors in each reconfigurable motion execution unit so that the reconfigurable motion execution units with the highest priority to the lowest priority perform angular motion step by step until the included angle of all adjacent rods is 180° and all rods are located in the motion reference direction; wherein the motion reference direction is any movable direction of the reconfigurable motion execution unit;

[0024] From the reconfigurable motion execution unit at the end of the robot mechanical body to the reconfigurable motion execution unit at the base of the robot mechanical body, each level of the reconfigurable motion execution unit moves independently, and when the real-time data sent back by the sensor does not change but the motor current increases, the angle range of the current reconfigurable motion execution unit is recorded;

[0025] The motor of the current reconfigurable motion execution unit is driven to move in the opposite direction and return to the initial position, and the initial angle range of each reconfigurable motion execution unit is obtained in order from low to high priority.

[0026] According to a robot reconfiguration method based on multi-sensor fusion provided by the present invention, the calculation of actual motion parameters specifically includes:

[0027] Calculate the actual angle value of each reconfigurable motion execution unit according to the angle sensor data;

[0028] Calculate the angular velocity of each reconfigurable motion execution unit according to the gyroscope data;

[0029] The position parameters of each reconfigurable motion execution unit are calculated based on the angle value and the angular velocity.

[0030] According to a robot reconfiguration method based on multi-sensor fusion provided by the present invention, step S5 specifically includes:

[0031] S51, establishing a DH parameter matrix according to the connection relationship and motion characteristics of the reconfigurable motion execution unit;

[0032] S52, establishing a homogeneous transformation matrix between adjacent reconfigurable motion execution units based on the DH parameter matrix;

[0033] S53, calculating the cumulative error value of the actual motion parameter relative to the simulation value, and when the cumulative error value exceeds a preset threshold, determining that the robot mechanical body has a fault;

[0034] S54, taking the faulty reconfigurable motion execution unit as the equivalent rod length, modifying the DH parameters and reconstructing the kinematic model;

[0035] S55. Re-plan the motion trajectory based on the reconstructed kinematic model. The robot mechanical body moves according to the re-planned motion trajectory. Repeat step S4, continue to obtain real-time monitoring data from the angle sensor and gyroscope, and calculate the actual motion parameters. Repeat steps S53-S55 until the robot mechanical body completes motion control.

[0036] The present invention also provides a robot reconfigurable system based on multi-sensor fusion, which adopts the robot reconfigurable method as described above, and comprises:

[0037] A structural division unit, used for dividing the robot mechanical body into a plurality of reconfigurable motion execution units, each of which comprises a joint, a motor, an angle sensor arranged at the input side of the motor, and a gyroscope arranged at the output side of the motor, wherein the angle sensor is concentrically connected with the motor, and a measuring axis of the gyroscope is tangent to the output side of the motor;

[0038] A calibration unit, used to calibrate the connection order and connection mode of any reconfigurable motion execution unit using a random sampling method to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model;

[0039] An equivalent mathematical model building unit is used to perform a paving operation on the optimization model in multiple movable directions to obtain an equivalent mathematical model;

[0040] A real-time monitoring unit is used to obtain real-time monitoring data of the angle sensor and the gyroscope and calculate the actual motion parameters, and input the real-time monitoring data into an equivalent mathematical model to obtain a simulation value;

[0041] The trajectory adjustment unit is used to make fault judgment on the robot mechanical body based on the actual motion parameters and simulation values. When the robot mechanical body fails, the parameters of the equivalent mathematical model are updated and the motion trajectory is replanned. The real-time monitoring and trajectory planning are repeated until the robot mechanical body completes motion control.

[0042] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any one of the above-mentioned robot reconfiguration methods is implemented.

[0043] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the robot reconfiguration method as described in any one of the above is implemented.

[0044] The present invention provides a robot reconfiguration method based on multi-sensor fusion, which realizes the reconfigurability and adaptability of the robot mechanical body by adding sensors to the motor to form a reconfigurable motion execution unit and establishing a corresponding equivalent mathematical model, so that the robot mechanical body can reconfigure its own form and function according to different environments and task requirements. At the same time, when a module fails, the entire system can eliminate the faulty module through reconfiguration to maintain the continuity of operation, and realizes real-time comparison and adjustment of simulated solutions and actual solutions through digital twin technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 is a flow chart of the robot reconfiguration method provided by the present invention;

[0047] Figure 2 It is a schematic diagram of the arrangement of sensors of the robot reconfiguration method provided by the present invention;

[0048] Figure 3 It is a schematic diagram of the minimum unit acceleration ratio in the robot mechanical body parameter calibration process of the robot reconfigurable method provided by the present invention;

[0049] Figure 4 It is a schematic diagram of the principle of the robot reconfiguration method provided by the present invention to constrain the mechanical motion of the system into the rotational motion of the reconfigurable motion execution unit;

[0050] Figure 5 It is a schematic diagram of the principle of the robot reconfiguration method provided by the present invention to constrain the mechanical motion of the system into the linear motion of the reconfigurable motion execution unit;

[0051] Figure 6 It is a flowchart of the robot reconfiguration method provided by the present invention;

[0052] Figure 7 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in combination with the drawings and embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] like Figure 1 As shown, the present invention provides a robot reconfiguration method based on multi-sensor fusion, comprising the following steps:

[0055] S1. Divide the robot mechanical body into several reconfigurable motion execution units, each reconfigurable motion execution unit includes a joint, a motor, an angle sensor arranged on the motor input side and a gyroscope arranged on the motor output side, wherein the angle sensor is concentrically connected to the motor, and the measuring axis of the gyroscope is tangent to the motor output side; wherein the robot mechanical body is divided according to the drive, the reducer, the connection structure of each output part in the robot mechanical body, the components of the connection structure, each end-effector in the robot mechanical body and the part of the end-effector, and the two adjacent reconfigurable motion execution units are connected by the connection structure.

[0056] like Figure 2 As shown, the angle sensor 1 is arranged at the tail of the input side motor 3 to measure the motor rotation angle. The angle sensor 1 should be connected concentrically with the device under test to improve the measurement accuracy and service life. The gyroscope 2 is arranged at the output side reducer 4 and has a measuring axis tangent to the output rotation to ensure that more accurate position information is obtained, providing safe data for the posture solution of the reconfigurable motion execution unit of the robot mechanical body. During the installation process, the zero point calibration and reference setting of the sensor should be correct to obtain accurate data. The sensor layout should ensure that the measured data can fully and accurately reflect the dynamic characteristics of the structure, have a high signal-to-noise ratio, and be sensitive to changes in the dynamic and static force parameters of the actual structure or environmental changes.

[0057] It can be understood that the base of the robot mechanical body is the first-level reconfigurable motion execution unit, which is also the highest-level reconfigurable motion execution unit. The sensor on the lowest-level reconfigurable motion execution unit rotates around the central axis. a The rotation speed of the higher-level reconfigurable motion execution unit is a b Satisfy a certain quantitative relationship, the relationship is as follows:

[0058]

[0059] Wherein, L represents the length of the rod connecting the two reconfigurable motion execution units, L1 represents the distance between the sensor on the lowest level reconfigurable motion execution unit and the central axis, L2 represents the distance between the sensor on the higher level reconfigurable motion execution unit and the central axis, θ represents the rotation angle of the reconfigurable motion execution unit, t0 represents the time when the reconfigurable motion execution unit starts to move, and t n Indicates the current moment.

[0060] By adding sensors to the motor to form a reconfigurable motion execution unit, real-time monitoring and data collection of the motion state of the robot mechanical body are realized, so that the robot mechanical body can accurately obtain the angle, speed and acceleration information of each joint, providing a reliable data basis for the precise control and fault diagnosis of the robot mechanical body, and also providing hardware support for the reconfigurability of the robot mechanical body.

[0061] In one embodiment of the present invention, the reconfigurable motion execution unit of the robot mechanical body includes a joint component, a reduction device, a FOC motor, a posture sensor mpu6050 and an angle sensor as5600.

[0062] S2. Use a random sampling method to calibrate the connection order and connection mode of any reconfigurable motion execution unit to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model.

[0063] Specifically, step S2 specifically includes:

[0064] S21, collecting data of the angle sensor in each reconfigurable motion execution unit, constraining the spatial structural motion of the robot mechanical body to the motion of the reconfigurable motion execution unit, and confirming the priority of each reconfigurable motion execution unit in the robot mechanical body, and using the least squares method to process the data of the angle sensor to obtain an initial model; wherein the motion of the reconfigurable motion execution unit includes the linear motion of the reconfigurable motion execution unit or the rotational motion of the reconfigurable motion execution unit;

[0065] S22, calibrating step by step from the end of the mechanism of the robot mechanical body toward the base direction until the reconfigurable motion execution units of all degrees of freedom in the robot mechanical body are calibrated to obtain an optimized model of the robot mechanical body;

[0066] S23, determining the initial angle range of two adjacent reconfigurable motion execution units and the length of the rod connecting the two adjacent reconfigurable motion execution units according to the structural constraints of the robot mechanical body;

[0067] S24, determining an optimization model according to the initial angle range and the length of a rod connecting two adjacent reconfigurable motion execution units.

[0068] like Figure 3 As shown in the figure, the horizontal axis represents time, the vertical axis represents the acceleration of the reconfigurable motion execution unit, and three different broken lines represent the acceleration changes of three different reconfigurable motion execution units, among which the red broken line represents the acceleration change of the reconfigurable motion execution unit with the highest priority, the yellow broken line represents the acceleration change of the reconfigurable motion execution unit with the second priority, and the blue broken line represents the acceleration change of the reconfigurable motion execution unit with the lowest priority. Specifically, the robot mechanical body is a 6-DOF manipulator as an example for explanation:

[0069] A set of reconfigurable motion execution units D = {D1, D2, D3, D4, D5, D6} of the internal mechanical body of the robot is arbitrarily selected, where D1 represents a reconfigurable motion execution unit of a first degree of freedom, D2 represents a reconfigurable motion execution unit of a second degree of freedom, D3 represents a reconfigurable motion execution unit of a third degree of freedom, D4 represents a reconfigurable motion execution unit of a fourth degree of freedom, D5 represents a reconfigurable motion execution unit of a fifth degree of freedom, and D6 represents a reconfigurable motion execution unit of a sixth degree of freedom;

[0070] Randomly select any set D from the robot's mechanical body i ={D a , D b , D c}, D i represents the geometry of the i-th reconfigurable motion execution unit, D a represents the reconfigurable motion execution unit with the ath degree of freedom, D b represents the reconfigurable motion execution unit of the bth degree of freedom, D c Reconfigurable motion execution unit represents the c-th degree of freedom. A reconfigurable motion execution unit with a degree of freedom is randomly selected to move. The angular acceleration sent back by the angle sensor is divided and calculated as follows:

[0071]

[0072] in, Degree of freedom D a The angular sensor sends back the angular acceleration. Degree of freedom D b The angular sensor transmits back the angular acceleration;

[0073] If the ratio k is greater than one, it indicates that the reconfigurable motion execution unit represented by the numerator has a lower priority than the reconfigurable motion execution unit represented by the denominator. If the sensor data of a reconfigurable motion execution unit is 0, it indicates that the reconfigurable motion execution unit has the highest priority in the set.

[0074] Reconfigurable motion execution unit D with residual degrees of freedom m ={D d , D e , D f}, D m represents the mth reconfigurable motion execution unit set, D d Denotes the reconfigurable motion execution unit with the dth degree of freedom, D e represents the reconfigurable motion execution unit of the e-th degree of freedom, D f The reconfigurable motion execution unit representing the f-th degree of freedom can also be used to infer the priority of the reconfigurable motion execution units in the set. If the priorities of all reconfigurable motion execution units cannot be distinguished, this method can be used to perform sampling and detection again until the priorities of all reconfigurable motion execution units in each system are distinguished.

[0075] like Figure 4 As shown, the two reconfigurable motion execution units can be connected by a straight rod or a curved rod. The connection structure described in the present invention specifically refers to the connecting rod between the executable units, and the connecting rod can be a connecting rod of any shape. The equivalent straight rod length between the executable units is derived from the calibration process of the rod length, and the angle range between the minimum units is derived from the paving operation referred to in the present invention. Figure 4 The principle of constraining the mechanical motion of the robot body to the rotational motion of the reconfigurable motion execution unit is:

[0076] The spatial structural motion of the robot's mechanical body is constrained to the rotational motion of the reconfigurable motion execution unit, so that the sensor on the reconfigurable motion execution unit rotates around the central axis. There is a specific quantitative relationship between the rotation speed of the sensor and the rotation speed of the higher-level reconfigurable motion execution unit. This motion constraint method can not only complete the system's rod length calibration, but also realize the measurement of the inherent parameters of the robot's mechanical body, providing necessary basic data support for subsequent system reconstruction and control.

[0077] like Figure 5 As shown, N is a reconfigurable motion execution unit, N+1 is a reconfigurable motion execution unit of the next level, and the angle sensor 1 and the posture sensor 5 are located on the same central axis. The principle of constraining the mechanical motion of the robot body to the linear motion of the reconfigurable motion execution unit is:

[0078] By constraining the mechanical motion of the robot's mechanical body to the linear motion of the reconfigurable motion execution unit, this motion constraint method can not only complete the system's rod length calibration work, but also have a specific impact on the end contact of the robotic arm. By analyzing this impact, we can better understand and control the system's motion characteristics, providing an important reference basis for the system's precise control and reconstruction.

[0079] Furthermore, the step S22 of calibrating step by step from the end of the mechanism of the robot mechanical body toward the base direction includes:

[0080] Step 1, recording the position of the reconfigurable motion execution unit to be calibrated in the initial model, and recording it as the zero point position;

[0081] Step 2: Control the reconfigurable motion execution unit to be calibrated to perform reverse growth motion until the return data of the sensor in the reconfigurable motion execution unit is 0, and the reconfigurable motion execution unit of the degree of freedom to be calibrated reaches the actual motion maximum position, and the actual motion maximum position is recorded;

[0082] Step 3. Record the actual maximum motion position of the reconfigurable motion execution unit of the degree of freedom to be calibrated, determine the midpoint position between the actual maximum motion position and the zero point position, perform a binary search between the midpoint position and the mechanical limit position of the reconfigurable motion execution unit in the robot mechanical body, and obtain the accurate position of the reconfigurable motion execution unit of the degree of freedom to be calibrated.

[0083] In one embodiment of the present invention, when calibrating the reconfigurable motion execution unit, the initial positions of all reconfigurable motion execution units are recorded as the zero position, the unit with the least interference with other reconfigurable motion execution units is selected for calibration, and the calibrated reconfigurable motion execution unit is placed at the position with the least interference with the lower-level reconfigurable motion execution unit. The above steps are repeated until the calibration of all reconfigurable motion execution units is completed, and the maximum angle range of each reconfigurable motion execution unit is obtained.

[0084] After the robot mechanical body determines the priority of each reconfigurable motion execution unit and reads the inherent parameters, the reconfigurable motion execution unit with the lowest priority starts to return to the zero position one by one. During this process, the rod length can be calibrated again according to the above method to obtain the optimized model.

[0085] S3. Perform paving operations on the optimized model in multiple movable directions to obtain an equivalent mathematical model.

[0086] The paving operation specifically includes:

[0087] Taking the base reconfigurable motion execution unit of the robot mechanical body as the starting point, setting the motion reference direction, driving the motors in each reconfigurable motion execution unit so that the reconfigurable motion execution units with the highest priority to the lowest priority perform angular motion step by step until the included angle of all adjacent rods is 180° and all rods are located in the motion reference direction; wherein the motion reference direction is any movable direction of the reconfigurable motion execution unit;

[0088] From the reconfigurable motion execution unit at the end of the robot mechanical body to the reconfigurable motion execution unit at the base of the robot mechanical body, each level of the reconfigurable motion execution unit moves independently, and when the real-time data sent back by the sensor does not change but the motor current increases, the angle range of the current reconfigurable motion execution unit is recorded;

[0089] The motor of the current reconfigurable motion execution unit is driven to move in the opposite direction and return to the initial position, and the initial angle range of each reconfigurable motion execution unit is obtained in order from low to high priority.

[0090] In the embodiment of the present invention, the motion reference direction may be a vertical direction or a horizontal direction, and the present application does not make any specific limitation on this.

[0091] As you can understand, the tiling operation is used to calibrate the limit conditions of the connection structure. If the motor current is abnormally increased, but the gyroscope reading does not fluctuate significantly, indicating that the change is not caused by external torque, but by internal factors, then it is determined whether the joint has reached the maximum angle value. The principle of determining whether the maximum angle value has been reached is as follows:

[0092] To ensure equipment safety and optimize performance, progressive load testing is required. By gradually adjusting the load and closely monitoring changes in motor current, the critical point of safe operation can be determined to effectively prevent damage to the motor equipment.

[0093] Adjust the mechanical joint to a safe state to avoid damage to the motor due to overload. Gradually reduce the load until the motor current drops to the normal operating range. Subsequently, gradually increase the load while closely monitoring the motor current to accurately identify the threshold at which the current begins to rise significantly. In this process, the load is gradually increased in a program-controlled manner to ensure the accuracy of the test. As the load gradually increases, the motor current will gradually rise. Record the exact moment when the current begins to increase significantly, which usually indicates that the motor is approaching its load limit. The joint phase referred to in the present invention includes the limit range of the motion angle of the reconfigurable motion execution unit in a predetermined coordinate system and the angle when the system end carries the maximum weight value of the load. The specific measurement method is to return the reconfigurable motion execution unit to the zero position, start from the end reconfigurable motion execution unit, evaluate the angle and load capacity of the reconfigurable motion execution unit according to the above-mentioned progressive load test method, and return to zero after the measurement is completed. Perform such measurement on the previous level reconfigurable motion execution unit to obtain the maximum load and angle motion data of the system reconfigurable motion execution unit. The relationship between the three-phase motor current and the load is as follows:

[0094]

[0095] Among them, I represents the three-phase motor current, U represents the voltage, and P represents the power.

[0096] Generally, the power factor of a three-phase motor is cosθ=0.85 and the efficiency η=0.85.

[0097] This progressive load testing method can not only effectively prevent motor overload, but also collect performance data about the motor under different working conditions. These data are of great value for understanding the working characteristics of the motor, optimizing design parameters, and improving system reliability. In addition, this method helps to identify and solve problems that may cause abnormal increase in current, such as mechanical friction, internal damage to the motor, or defects in the control algorithm, thereby ensuring the stable operation and long-term performance of the robot's mechanical body.

[0098] In one embodiment of the present invention, there are four reconfigurable motion execution units in the optimization model, and they are arranged in a trapezoidal shape. First, the bottommost base reconfigurable motion execution unit (recorded as unit No. 1) is taken as the starting point, and the vertical upward direction is taken as the reference direction. The motors of each unit are driven in sequence from high to low priority (i.e., from No. 1 to No. 4): first, the motor of unit No. 1 is driven to align it with the vertical direction, and then the motor of unit No. 2 is driven to adjust the angle between it and the connecting rod of unit No. 1 to 180 degrees, and then the motor of unit No. 3 is driven to adjust the angle between it and the connecting rod of unit No. 2 to 180 degrees, and finally the motor of unit No. 4 is driven to adjust the angle between it and the connecting rod of unit No. 3 to 180 degrees. At this time, the connecting rods of all units are in the vertical direction. direction; after completing the vertical alignment, start testing the angular range of each unit in turn from unit No. 4 at the end toward unit No. 1 on the base, drive the motor of each unit to make it move individually, and monitor the sensor data and motor current at the same time. When it is observed that the angle and posture data sent back by the sensor in real time remain unchanged while the motor current increases significantly, record the angular position of the current unit as the limit of its motion range; after completing the angular range test of each unit, drive the motor of the unit to move in the opposite direction to return it to its initial position, and obtain and record the initial angular range of each unit in the trapezoidal structure in order from No. 4 to No. 1 (i.e., priority from low to high), and finally complete the angular range calibration process of the entire paving operation.

[0099] In one embodiment of the present invention, after the optimization model is flattened, the joint angle movement range must be obtained to make the angle range of the equivalent mathematical model and the angle range of the robot mechanical body more accurate. The method for obtaining the joint angle movement range is the same as the principle of determining whether the maximum angle value has been reached, and will not be repeated here.

[0100] S4, obtaining real-time monitoring data of the angle sensor and the gyroscope and calculating actual motion parameters, and inputting the real-time monitoring data into an equivalent mathematical model to obtain simulation values; wherein the actual motion parameters include the actual angle value, angular velocity and position parameters of the reconfigurable motion execution unit.

[0101] Specifically, the calculating of the actual motion parameters includes:

[0102] Calculate the actual angle value of each reconfigurable motion execution unit according to the angle sensor data;

[0103] Calculate the angular velocity of each reconfigurable motion execution unit according to the gyroscope data;

[0104] The position parameters of each reconfigurable motion execution unit are calculated based on the angle value and the angular velocity.

[0105] S5. Perform fault judgment on the robot mechanical body based on the actual motion parameters and simulation values. When the robot mechanical body fails, update the parameters of the equivalent mathematical model and re-plan the motion trajectory, and repeat steps S4-S5 until the robot mechanical body completes motion control.

[0106] Specifically, step S5 specifically includes:

[0107] S51, establishing a DH parameter matrix according to the connection relationship and motion characteristics of the reconfigurable motion execution unit;

[0108] S52, establishing a homogeneous transformation matrix between adjacent reconfigurable motion execution units based on the DH parameter matrix;

[0109] S53, calculating the cumulative error value of the actual motion parameter relative to the simulation value, and when the cumulative error value exceeds a preset threshold, determining that the robot mechanical body has a fault;

[0110] S54, taking the faulty reconfigurable motion execution unit as the equivalent rod length, modifying the DH parameter matrix and reconstructing the kinematic model;

[0111] S55. Re-plan the motion trajectory based on the reconstructed kinematic model. The robot mechanical body moves according to the re-planned motion trajectory. Repeat step S4, continue to obtain real-time monitoring data from the angle sensor and gyroscope, and calculate the actual motion parameters. Repeat steps S53-S55 until the robot mechanical body completes motion control.

[0112] The present invention realizes the continuity and smoothness of the robot's mechanical body movement by optimizing the planned path and calculating the real-time error accumulation, combining obstacle position judgment with a data inheritance mechanism. At the same time, it can effectively avoid obstacles during the movement, and improves the operating efficiency of the robot's mechanical body through the data inheritance mechanism, reducing the time for repeated calculation and calibration.

[0113] It can be understood that when the current reconfigurable motion execution unit fails, the current reconfigurable motion execution unit is locked, that is, the current reconfigurable motion execution unit is recorded as connecting rod data into the previous level reconfigurable motion execution unit to form a reconfigurable motion execution unit.

[0114] In one embodiment of the present invention, by establishing a forward kinematic model of a reconfigurable motion execution unit, based on the real-time data collected by the angle sensor and the gyroscope, combined with the known geometric dimensions of the connecting rod, the position and posture of the end of the robot mechanical body are calculated using the homogeneous transformation matrix established by the DH parameters. On this basis, through the inverse kinematics solution process, according to the desired end position and posture and the coordinate transformation relationship calibrated by the DH parameters, the motion parameters of each reconfigurable motion execution unit required to achieve the target motion are calculated, and each motor is driven to perform the corresponding motion accordingly; when the system detects a fault, the faulty reconfigurable motion execution unit is used as the equivalent rod length and the DH parameters and coordinate transformation relationship are re-established, and the kinematic equations are re-solved with reduced degrees of freedom, thereby realizing trajectory planning and motion control under the fault state, ensuring that the robot mechanical body can still maintain basic functions after the fault occurs.

[0115] The homogeneous transformation matrix of the robot mechanical body coordinate system {j} relative to the upper level coordinate system {j-1} is:

[0116]

[0117] Among them, j represents the minimum unit coordinate system number, θ j Represents the joint rotation angle, which is x j-1 Along z j Turn to x j The directed angle of the triangle, C represents the abbreviation of the trigonometric function sine, S represents the abbreviation of the trigonometric function cosine, a j represents the length of a single link in a single link joint, that is, the length of the common normal from the coordinate system {j} to {j+1}, α j Represents the coordinate axis z j With z j+1 The deflection angle between the connecting rod and the connecting rod (i.e. the connecting rod bending angle) is positive if it is counterclockwise, d j represents the connecting rod offset, which is x j-1 With z j The intersection point of the link j coordinate system along z j Directed distance in direction, represents the joint rotation angle θ j The cosine value of Indicates the connecting rod bending angle α j The cosine value of represents the joint rotation angle θ j The sine value of Represents the homogeneous transformation matrix from the j-1th coordinate system to the jth coordinate system.

[0118] In the process of solving the motion of the machine system, the position and posture of the end of the robot mechanical body can be obtained by the coordinate system transformation of the connecting rod. The key to establishing an equivalent mathematical model is to find the mapping relationship between the local coordinate system and the global coordinate system. The transformation matrix of the adjacent connecting rods can be used to establish the position and posture matrix of the end of the robot mechanical body based on the above coordinate system and DH model parameters. The established forward kinematics formula of the robot mechanical body is as follows:

[0119] 0 A T = 0 A1 1 A2 2 A3 3 A4 4 A5 5 A6

[0120] in, 0 A T represents the overall transformation matrix from the base coordinate system {0} to the end pose coordinate system {T}, 0 A1 represents the transformation matrix from the base coordinate system {0} to the first link coordinate system {1}, 1 A2 represents the transformation matrix from the first link coordinate system {1} to the second link coordinate system {2}, 2 A3 represents the transformation matrix from the second link coordinate system {2} to the third link coordinate system {3}, 3 A4 represents the transformation matrix from the third link coordinate system {3} to the fourth link coordinate system {4}, 4 A5 represents the transformation matrix from the 4th link coordinate system {4} to the 5th link coordinate system {5}, 5 A6 represents the transformation matrix from the fifth link coordinate system {5} to the sixth link coordinate system {6}.

[0121] Furthermore, the equivalent angle range of the reconfigurable motion execution unit is calculated based on its structure before locking. The angle between the two rods connected to the reconfigurable motion execution unit can be known through the motion data of the current reconfigurable motion execution unit before locking, and the angle constraint range of the reconfigurable motion execution unit on the previous level can be inferred through the cosine theorem when the calibrated rod length information is known, so as to obtain the equivalent motion angle range. The specific mathematical formula is as follows:

[0122]

[0123] Wherein, r represents the length of the connecting rod between the current reconfigurable motion execution unit and the previous reconfigurable motion execution unit, h represents the length of the connecting rod between the current reconfigurable motion execution unit and the next reconfigurable motion execution unit, and g represents the distance between r and h;

[0124] For the reconfigurable motion execution unit that is locked at the previous level, ∠A is the limitation caused by the equivalent rod length generated by the reconfigurable motion execution unit failure. By finding the size of ∠A, the angle limitation problem of the reconfigurable motion execution unit in the system can be solved, and the solution of the reconfigurable motion execution unit within this angle range can be eliminated in the process of generating simulation solutions for the parameter matrix.

[0125] In one embodiment of the present invention, by establishing an operating parameter system of a reconfigurable motion execution unit, after confirming the role of each reconfigurable motion execution unit on the overall robot mechanical body based on the monitoring data of the reconfigurable motion execution unit, the robot mechanical body judges the degree of freedom. If a faulty reconfigurable motion execution unit is found according to the above parameters, the parameter matrix is ​​automatically modified to generate an actual parameter matrix.

[0126] Specifically, in order to maintain stability and data accuracy, it is necessary to obtain the sequential relationship between the joints through random selection, and then calculate the various parameters of the robot's mechanical body starting from the end joint of the robot's mechanical body. After all joints are calculated, in order to ensure the integrity of the data and check the data, it is necessary to traverse again from the initial end joint to the end joint.

[0127] In one embodiment of the present invention, when a robot mechanical body fails, an equivalent mathematical model is required to re-plan the motion trajectory to ensure that the reconstructed robot mechanical body can efficiently and stably complete the predetermined motion task. The re-planning of the motion trajectory includes two stages: in the coarse interpolation stage, the system first performs global random sampling of the motion trajectory of each reconfigurable motion execution unit of the reconstructed robot mechanical body, and constructs a local sampling area based on the validity of the sampling data to obtain the key trajectory points required to complete the predetermined task. These trajectory points correspond to the motion solution of the reconfigurable motion execution unit calculated based on the DH parameter matrix; in the fine interpolation stage, considering the requirements of the reconstructed robot mechanical body for motion time and accuracy, the system uses a quintic polynomial interpolation method to smooth the trajectory points obtained by coarse interpolation, where the quintic polynomial interpolation formula is:

[0128] θ(t)=a0+a1t+a2t2+a3t3+a4t4+a5t5

[0129] Wherein, θ(t) represents the angular position of the reconfigurable motion execution unit at time t, a0 represents the constant term coefficient of the initial angular position, a1 represents the first-order term coefficient of the initial angular velocity, a2 represents the second-order term coefficient of the initial angular acceleration, a3 represents the third-order term coefficient of the angular jerk, a4 represents the fourth-order term coefficient of the angular jerk, a5 represents the fifth-order term coefficient of the angular jerk, t represents time, t2 represents the square term of time, t3 represents the cube term of time, t4 represents the fourth-order term of time, and t5 represents the fifth-order term of time;

[0130] By combining global sampling and local optimization, the integrity of trajectory planning is guaranteed while local precise control is achieved. However, it should be noted that the randomness in the sampling process may lead to uncertainty in the sampling area. Therefore, in practical applications, it is necessary to adjust the sampling strategy and optimize the parameters to improve the sampling efficiency and stability of trajectory optimization.

[0131] In one embodiment of the present invention, during the movement of the robot arm, due to the limitation of the spatial structure, some system structures may not be able to move according to the originally planned path data. The present invention accumulates the difference between the data including angular velocity, angle, position, etc. transmitted back by the sensors configured by all reconfigurable motion execution units and the simulation results. If the accumulated error of a certain reconfigurable motion execution unit is greater than the threshold, it is judged that the robot arm encounters an obstacle during the movement, and the approximate position of the obstacle can be judged by the movement angle of the reconfigurable motion execution unit at this time.

[0132] The specific judgment method is to perform a second integration of the angular acceleration sent back by the sensor of the reconfigurable motion execution unit to obtain the angle between the obstacle and the zero position of the reconfigurable motion execution unit:

[0133]

[0134] Among them, θ d represents the angle between the obstacle and the zero position of the reconfigurable motion execution unit, a t Represents the angular acceleration sent back by the sensor, d t represents the time differential term;

[0135] In the process of the robot's mechanical body avoiding obstacles, the approximate position of the obstacle is known, from which the geometric restriction relationship of the obstacle in the horizontal and vertical directions can be obtained, and data inheritance is prepared for the next movement. The actual system parameters calculated from the previous system movement can be directly inherited to participate in the next movement, without having to perform operations such as determining the rod length and the center of gravity position through the minimum unit rotation again, thereby saving time and improving efficiency.

[0136] The present invention forms a reconfigurable motion execution unit by adding sensors to the motor and establishes a corresponding equivalent mathematical model, thereby realizing the reconfigurability and adaptability of the robot mechanical body, enabling the robot mechanical body to reconfigure its own form and function according to different environments and task requirements. At the same time, when a module fails, the entire system can eliminate the faulty module through reconfiguration to maintain the continuity of operation, and realizes real-time comparison and adjustment of simulated solutions and actual solutions through digital twin technology.

[0137] like Figure 6As shown, specifically, the robot reconfiguration method is described with a specific embodiment:

[0138] First, the touch screen display system waits to receive motion instructions, obtains the structural information of the robot's mechanical body, and divides it into several reconfigurable motion execution units. The reconfigurable motion execution units are calibrated using a random sampling method to obtain an optimization model, and the reconfigurable motion execution units are further calibrated based on the priority order of the reconfigurable motion execution units in the optimization model to obtain an optimization model; the paving operation of the movable direction is performed based on the optimization model to obtain an equivalent mathematical model; if the equivalent mathematical model accepts the motion instructions of the robot's mechanical body, stream data processing is performed, that is, the actual motion parameters of the robot's mechanical body are calculated, posture settlement is performed based on the actual motion parameters, and a DH parameter matrix is ​​generated. The settlement result is transmitted to the kinematic model to determine whether the simulated value is the same as the actual value. If not, it is determined that the equivalent mathematical model has an external equivalent solution, and the equivalent mathematical model re-plans the trajectory, optimizes the trajectory, drives the motor of the robot's mechanical body to move according to the optimized trajectory, and calculates the accumulated error between the actual motion parameters and the simulated value of each reconfigurable motion execution unit in real time. The motion state is continuously monitored through a small disturbance feedback mechanism. When the error exceeds the preset threshold, the robot mechanical body will determine that the corresponding unit is faulty, and reconstruct the equivalent mathematical model by taking the faulty unit as the equivalent rod length, re-determine the motor position and perform angle calibration. The equivalent mathematical model performs trajectory planning and optimizes the motion trajectory by combining coarse interpolation and fine interpolation. The coarse interpolation obtains key trajectory points through global sampling, and the fine interpolation uses a quintic polynomial interpolation method to achieve trajectory smoothing. During the execution of the movement, the robot mechanical body continuously judges and records path obstacles, and saves the obstacle information through a data inheritance mechanism and uses it for subsequent motion planning. The robot mechanical body drives the motor to execute the planned movement through a PWM signal, and transmits the real-time motion parameters of each reconfigurable motion execution unit back to the control system. By comparing the actual motion data with the calculation results of the equivalent mathematical model, the system operation status is continuously monitored. At the same time, the operation status is displayed on the touch screen in real time and new control instructions are received, thereby realizing precise control of the robot mechanical body in normal and fault states.

[0139] Among them, in the reading stage of streaming data, the present invention is based on CAN bus communication, and through decision-making, three-layer instructions between central control and minimum control, are interpreted by specific interpreters between each other, and data is sent down layer by layer to achieve collaborative work of the robot mechanical body.

[0140] The present invention provides a robot reconfigurable system based on multi-sensor fusion, which adopts the robot reconfigurable method as described above, and comprises:

[0141] A structural division unit, used for dividing the robot mechanical body into a plurality of reconfigurable motion execution units, each of which comprises a joint, a motor, an angle sensor arranged at the input side of the motor, and a gyroscope arranged at the output side of the motor, wherein the angle sensor is concentrically connected with the motor, and a measuring axis of the gyroscope is tangent to the output side of the motor;

[0142] A calibration unit, used to calibrate the connection order and connection mode of any reconfigurable motion execution unit using a random sampling method to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model;

[0143] An equivalent mathematical model building unit is used to perform a paving operation on the optimization model in multiple movable directions to obtain an equivalent mathematical model;

[0144] A real-time monitoring unit is used to obtain real-time monitoring data of the angle sensor and the gyroscope and calculate the actual motion parameters, and input the real-time monitoring data into an equivalent mathematical model to obtain a simulation value;

[0145] The trajectory adjustment unit is used to make fault judgment on the robot mechanical body based on the actual motion parameters and simulation values. When the robot mechanical body fails, the parameters of the equivalent mathematical model are updated and the motion trajectory is replanned. The real-time monitoring and trajectory planning are repeated until the robot mechanical body completes motion control.

[0146] The robot reconfigurable device provided by the present invention is described below. The robot reconfigurable device described below and the robot reconfigurable method described above can be referenced to each other.

[0147] Figure 7 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7As shown, the electronic device may include: a processor (processor) 710, a communication interface (CommunicationsInterface) 720, a memory (memory) 730 and a communication bus 740, wherein the processor 710, the communication interface 720, and the memory 730 communicate with each other through the communication bus 740. The processor 710 can call the logic instructions in the memory 730 to execute the robot reconfiguration method, which includes: dividing the robot mechanical body into a number of reconfigurable motion execution units; using a random sampling method to calibrate the connection order and connection method of any reconfigurable motion execution unit to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model; performing a paving operation on the optimized model in multiple movable directions to obtain an equivalent mathematical model; obtaining real-time monitoring data of the angle sensor and the gyroscope and calculating the actual motion parameters, and inputting the real-time monitoring data into the equivalent mathematical model to obtain a simulation value; performing a fault judgment on the robot mechanical body based on the actual motion parameters and the simulation values, and when the robot mechanical body fails, updating the parameters of the equivalent mathematical model and re-planning the motion trajectory, real-time monitoring and trajectory planning, until the robot mechanical body completes motion control.

[0148] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0149] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the robot reconfiguration method provided by the above methods, which includes: dividing the robot mechanical body into a number of reconfigurable motion execution units; using a random sampling method to calibrate the connection order and connection method of any reconfigurable motion execution unit to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model; performing a paving operation on the optimized model in multiple movable directions to obtain an equivalent mathematical model; obtaining real-time monitoring data of the angle sensor and the gyroscope and calculating the actual motion parameters, and inputting the real-time monitoring data into the equivalent mathematical model to obtain a simulation value; performing a fault judgment on the robot mechanical body based on the actual motion parameters and the simulation values, and when the robot mechanical body fails, updating the parameters of the equivalent mathematical model and re-planning the motion trajectory, real-time monitoring and trajectory planning, until the robot mechanical body completes motion control.

[0150] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the robot reconfiguration method provided by the above-mentioned methods, the method comprising: dividing the robot mechanical body into a number of reconfigurable motion execution units; using a random sampling method to calibrate the connection order and connection method of any reconfigurable motion execution unit to obtain an initial model, and calibrating the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model; performing a paving operation on the optimized model in multiple movable directions to obtain an equivalent mathematical model; obtaining real-time monitoring data of the angle sensor and the gyroscope and calculating the actual motion parameters, and inputting the real-time monitoring data into the equivalent mathematical model to obtain a simulation value; performing a fault judgment on the robot mechanical body based on the actual motion parameters and the simulation values, and when a fault occurs in the robot mechanical body, updating the parameters of the equivalent mathematical model and re-planning the motion trajectory, real-time monitoring and trajectory planning, until the robot mechanical body completes motion control.

[0151] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0152] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A robot reconfiguration method based on multi-sensor fusion, characterized in that: The following steps are involved: S1. Divide the robot mechanical body into a plurality of reconfigurable motion execution units, each of which includes a joint, a motor, an angle sensor disposed on the input side of the motor, and a gyroscope disposed on the output side of the motor, wherein the angle sensor is concentrically connected to the motor, and a measuring axis of the gyroscope is tangent to the output side of the motor; S2. Use a random sampling method to calibrate the connection order and connection mode of any reconfigurable motion execution unit to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model; S3, performing paving operations on the optimized model in multiple movable directions to obtain an equivalent mathematical model; S4, obtaining real-time monitoring data of the angle sensor and the gyroscope and calculating the actual motion parameters, and inputting the real-time monitoring data into an equivalent mathematical model to obtain a simulation value; S5. Perform fault judgment on the robot mechanical body based on the actual motion parameters and simulation values. When the robot mechanical body fails, update the parameters of the equivalent mathematical model and re-plan the motion trajectory, and repeat steps S4-S5 until the robot mechanical body completes motion control.

2. The robot reconfiguration method based on multi-sensor fusion according to claim 1, characterized in that: The robot mechanical body is divided according to the drive, reducer, connection structure of each output part in the robot mechanical body, components of the connection structure, each end effector in the robot mechanical body and parts of the end effector, and the two adjacent reconfigurable motion execution units are connected by a connection structure.

3. The robot reconfiguration method based on multi-sensor fusion according to claim 2 is characterized in that: Step S2 specifically includes: S21, collecting data of the angle sensor in each reconfigurable motion execution unit, constraining the spatial structural motion of the robot mechanical body to the motion of the reconfigurable motion execution unit, and confirming the priority of each reconfigurable motion execution unit in the robot mechanical body, using the least squares method to process the data of the angle sensor to obtain an initial model of the robot mechanical body; S22, calibrating step by step from the end of the mechanism of the robot mechanical body toward the base direction until the reconfigurable motion execution units of all degrees of freedom in the robot mechanical body are calibrated to obtain an optimized model of the robot mechanical body; S23, determining the initial angle range of two adjacent reconfigurable motion execution units and the length of the rod connecting the two adjacent reconfigurable motion execution units according to the structural constraints of the robot mechanical body; S24, determining an optimization model according to the initial angle range and the length of a rod connecting two adjacent reconfigurable motion execution units.

4. The robot reconfiguration method based on multi-sensor fusion according to claim 3 is characterized in that: The step S22 of calibrating step by step from the end of the mechanism of the robot mechanical body toward the base includes: Step 1, recording the position of the reconfigurable motion execution unit to be calibrated in the initial model, and recording it as the zero point position; Step 2: Control the reconfigurable motion execution unit to be calibrated to perform reverse growth motion until the return data of the sensor in the reconfigurable motion execution unit is 0, and the reconfigurable motion execution unit to be calibrated reaches the actual motion maximum position, and the actual motion maximum position is recorded; Step 3: Determine the midpoint between the actual maximum motion position and the zero point position, perform a binary search between the midpoint position and the mechanical limit position of the reconfigurable motion execution unit in the robot mechanical body, and obtain the exact position of the reconfigurable motion execution unit of the degree of freedom to be calibrated.

5. The robot reconfiguration method based on multi-sensor fusion according to claim 1, characterized in that: The paving operation specifically includes: Taking the base reconfigurable motion execution unit of the robot mechanical body as the starting point, setting the motion reference direction, driving the motors in each reconfigurable motion execution unit so that the reconfigurable motion execution units with the highest priority to the lowest priority perform angular motion step by step until the included angle of all adjacent rods is 180° and all rods are located in the motion reference direction; wherein the motion reference direction is any movable direction of the reconfigurable motion execution unit; From the reconfigurable motion execution unit at the end of the robot mechanical body to the reconfigurable motion execution unit at the base of the robot mechanical body, each level of the reconfigurable motion execution unit moves independently, and when the real-time data sent back by the sensor does not change but the motor current increases, the angle range of the current reconfigurable motion execution unit is recorded; The motor of the current reconfigurable motion execution unit is driven to move in the opposite direction and return to the initial position, and the initial angle range of each reconfigurable motion execution unit is obtained in order from low to high priority.

6. The robot reconfiguration method based on multi-sensor fusion according to claim 1, characterized in that: The calculating of the actual motion parameters specifically includes: Calculate the actual angle value of each reconfigurable motion execution unit according to the angle sensor data; Calculate the angular velocity of each reconfigurable motion execution unit according to the gyroscope data; The position parameters of each reconfigurable motion execution unit are calculated based on the angle value and the angular velocity.

7. The robot reconfiguration method based on multi-sensor fusion according to claim 1, characterized in that: Step S5 specifically includes: S51, establishing a DH parameter matrix according to the connection relationship and motion characteristics of the reconfigurable motion execution unit; S52, establishing a homogeneous transformation matrix between adjacent reconfigurable motion execution units based on the DH parameter matrix; S53, calculating the cumulative error value of the actual motion parameter relative to the simulation value, and when the cumulative error value exceeds a preset threshold, determining that the robot mechanical body has a fault; S54, taking the faulty reconfigurable motion execution unit as the equivalent rod length, modifying the DH parameter matrix and reconstructing the kinematic model; S55. Re-plan the motion trajectory based on the reconstructed kinematic model. The robot mechanical body moves according to the re-planned motion trajectory. Repeat step S4, continue to obtain real-time monitoring data from the angle sensor and gyroscope, and calculate the actual motion parameters. Repeat steps S53-S55 until the robot mechanical body completes motion control.

8. A robot reconfigurable system based on multi-sensor fusion, characterized in that: The robot reconfiguration method according to any one of claims 1 to 7 comprises: A structural division unit, used for dividing the robot mechanical body into a plurality of reconfigurable motion execution units, each of which comprises a joint, a motor, an angle sensor arranged at the input side of the motor, and a gyroscope arranged at the output side of the motor, wherein the angle sensor is concentrically connected with the motor, and a measuring axis of the gyroscope is tangent to the output side of the motor; A calibration unit, used to calibrate the connection order and connection mode of any reconfigurable motion execution unit using a random sampling method to obtain an initial model, and calibrate the connection structure between two adjacent reconfigurable motion execution units according to the priority information in the initial model to obtain an optimized model; An equivalent mathematical model building unit is used to perform a paving operation on the optimization model in multiple movable directions to obtain an equivalent mathematical model; A real-time monitoring unit is used to obtain real-time monitoring data of the angle sensor and the gyroscope and calculate the actual motion parameters, and input the real-time monitoring data into an equivalent mathematical model to obtain a simulation value; The trajectory adjustment unit is used to make fault judgment on the robot mechanical body based on the actual motion parameters and simulation values. When the robot mechanical body fails, the parameters of the equivalent mathematical model are updated and the motion trajectory is replanned. The real-time monitoring and trajectory planning are repeated until the robot mechanical body completes motion control.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the robot reconfiguration method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the robot reconfiguration method according to any one of claims 1 to 7 is implemented.

Images