Multi-degree-of-freedom motion robot system and control method

Abstract

The invention provides a multi-degree-of-freedom motion robot system and a control method. The multi-degree-of-freedom motion robot system comprises a robot upper body, a connecting mechanism and a movable chassis. A left-right motion module, a front-back motion module, a rotary motion module and a main controller are integrally arranged on the movable chassis; the left-right swing motion module and the front-back swing motion module are used for driving the robot upper body to achieve left-right swing motion and front-back pitching motion correspondingly. The rotary motion module is used for driving the whole robot to rotate around a central shaft; and the main controller is used for independently or cooperatively driving the three motion modules according to a control instruction, so that the robot executes multi-degree-of-freedom composite motion including pure swinging, pure pitching, swinging and pitching composite, in-situ rotation and swinging / pitching and rotation composite. By the adoption of the method, it is guaranteed that single motion is accurate and controllable, smooth coordination of composite motion is achieved, the motion expression forms of the robot are remarkably enriched, and the attraction and interaction effect of the robot in a target scene are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robots, in particular to a multi-degree-of-freedom robot system and a control method. BACKGROUND

[0002] With the rapid development of intelligent robot technology, its application in public scenarios such as commercial marketing and exhibition is increasingly widespread. The core requirement of such scenarios is to attract the attention of the crowd through the dynamic performance of the robot and to improve the interactive atmosphere of the scene, so the motion attraction of the robot becomes a key factor affecting the application effect.

[0003] In the prior art, the mainstream marketing and display robots mostly adopt a wheeled driving structure, and the core motion function is limited to movement in the plane, that is, through the rotation of the wheeled chassis to realize forward movement, backward movement, turning and other basic displacements, and no other motion mechanism capable of changing the posture is provided.

[0004] The core defect of the prior art is that the motion form is single and lacks dynamic change highlights. Since the robot can only complete simple displacement motion, the motion process is mechanical and monotonous, lacks sufficient visual impact, is difficult to continuously attract the attention of the crowd, cannot create an active interactive atmosphere, and thus cannot meet the core requirement of high attraction and strong topic in the marketing and display scenarios. SUMMARY

[0005] Therefore, the purpose of the present application is to provide a multi-degree-of-freedom robot system and a control method, which not only ensures the precise controllability of single motion, but also realizes the smooth coordination of compound motion, significantly enriches the motion performance form of the robot, and effectively improves its attraction and interactive effect in the target scene.

[0006] In a first aspect, the embodiments of the present application provide a multi-degree-of-freedom robot system, which comprises a robot upper body, a connecting mechanism and a mobile chassis; the mobile chassis is integrally provided with a left-right swinging motion module, a front-back swinging motion module, a rotating motion module and a main controller; The left-right swinging motion module and the front-back swinging motion module are connected to the robot upper body through the connecting mechanism, and are used to respectively drive the robot upper body to realize left-right swinging motion and front-back pitching motion; The rotating motion module is used to drive the whole robot to rotate around the central axis; The main controller is connected to the left-right swinging motion module, the front-back swinging motion module and the rotating motion module, and is used to independently or cooperatively drive the three motion modules according to a control instruction, so that the robot performs multi-degree-of-freedom compound motion including pure swinging, pure pitching, swinging and pitching compound, rotation in place, and swinging / pitching and rotation compound.

[0007] Optionally, the left-right swing movement module comprises a left-right sliding table and a first sliding table driver driving the same, the front-back swing movement module comprises a front-back sliding table and a second sliding table driver driving the same, and the rotating movement module comprises a wheel hub motor and a wheel hub driver driving the same; The left-right sliding table and the front-back sliding table are arranged in a cross slider structure in a top-bottom stacking manner and are centrally installed based on a center point of the robot; The connecting mechanism comprises a ball screw, and the first sliding table driver and the second sliding table driver are connected with corresponding ball screws through couplings respectively; A rotating part of the rotating movement module is an independent table surface located above the chassis, the cross slider structure is installed on the independent table surface, and the independent table surface rotates by relying on an output shaft of the wheel hub motor.

[0008] Optionally, the main controller comprises: An edge computing processor is configured to process perception data and interaction information, identify user intention and scene context, and generate high-level motion intention instructions containing emotional semantics and motion patterns based on the same; A motion control unit is in communication connection with the edge computing processor and is configured to receive the high-level motion intention instructions, calculate executable cooperative motion trajectories of the left-right sliding table, the front-back sliding table and the rotating platform based on the same, and run a synchronous control algorithm and a real-time closed-loop control to drive the first sliding table driver, the second sliding table driver and the wheel hub motor.

[0009] Optionally, the edge computing processor is connected with an interaction perception subsystem configured to provide multi-modal input for generating the high-level motion intention instructions and execute output feedback; The interaction perception subsystem comprises a camera configured to collect visual information, a microphone array configured to collect audio information, a display and a loudspeaker configured to output visual and audio feedback; The edge computing processor is configured to fuse and analyze the visual information and the audio information, dynamically generate or adjust the high-level motion intention instructions, and coordinate the output of the motion control unit, the display and the loudspeaker.

[0010] Optionally, the motion control unit is connected with a motion perception subsystem configured to provide high-precision feedback for the real-time closed-loop control; The motion perception subsystem comprises an attitude sensor configured to monitor real-time attitude of the robot body and a displacement sensor configured to monitor real-time positions of the left-right sliding table and the front-back sliding table; The motion control unit is configured to dynamically compensate and adjust in the process of calculating and executing the cooperative motion trajectories based on the attitude and position feedback data, so as to ensure the accuracy and smoothness of the compound motion.

[0011] Secondly, embodiments of this application provide a control method for a multi-degree-of-freedom motion robot system, applicable to the multi-degree-of-freedom motion robot system described in any optional embodiment of the first aspect above, the method comprising: Receive target motion instructions; Based on the target motion command, generate a coordinated motion trajectory for the left-right swaying motion module, the forward-backward swaying motion module, and the rotation motion module; The coordinated motion trajectory is smoothly planned to generate control commands for each motion module; The control commands used to drive the left and right swaying motion module, the forward and backward swaying motion module, and the rotation motion module are processed synchronously and in a coordinated manner. Output control commands and perform real-time closed-loop adjustments based on sensor feedback to execute the multi-degree-of-freedom composite motion.

[0012] Optionally, generating the coordinated motion trajectory for the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module according to the target motion command includes: The target motion command is parsed to generate a high-level motion intent command that includes motion patterns and emotional semantics; Based on the high-level motion intention command, the current robot state, and kinematic constraints, trajectory data for the spatiotemporal coordination of the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module is calculated and used as the coordinated motion trajectory.

[0013] Optionally, the step of smoothly planning the coordinated motion trajectory and generating control commands for each motion module includes: The S-shaped velocity curve algorithm is used to plan the trajectory segments corresponding to the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module in the cooperative motion trajectory. By constraining the jerk during the motion process, a smooth motion profile with continuously changing acceleration, velocity, and position is generated. Based on the smooth motion profile, the target position, target velocity, and target acceleration of the first slide driver of the left-right swaying motion module, the second slide driver of the front-back swaying motion module, and the hub driver of the rotational motion module are calculated in each control cycle, and used as control commands for each motion module.

[0014] Optionally, the step of synchronizing and coordinating the control commands used to drive the left-right swaying motion module and the forward-backward swaying motion module includes: Obtain the real-time motion status of the left-right swaying motion module and the forward-backward swaying motion module; Calculate the tracking error between the motion state of the two modules and their corresponding control commands; Calculate the synchronization error between the tracking errors of the two modules; A synchronization compensation amount is generated based on the synchronization error; The control commands for the left-right swaying motion module and the forward-backward swaying motion module are corrected using the synchronization compensation amount to suppress the synchronization error between the two modules during the motion process.

[0015] Optionally, the real-time closed-loop adjustment based on sensor feedback to execute the multi-degree-of-freedom composite motion includes: An enhanced PID control algorithm is used for closed-loop servo control, and the algorithm includes integral anti-saturation and differential filtering mechanisms. Alternatively, feedforward compensation based on a multi-dimensional compensation model can be used to offset errors caused by system nonlinearity and environmental disturbances.

[0016] The technical solution provided in this application includes, but is not limited to, the following beneficial effects: The mobile chassis integrates left-right swaying motion modules, forward-backward swaying motion modules, rotation motion modules, and a main controller, achieving a unified design with multiple motion functions. This layout eliminates the need for additional external motion mechanisms, simplifying the system structure, reducing equipment size and weight, and facilitating signal transmission and coordinated control between modules. It avoids the complex connections and response delays caused by distributed layouts, thus improving the stability and reliability of system operation.

[0017] The left-right swaying motion module and the forward-backward swaying motion module are independently connected to the robot's upper body via a connecting mechanism, respectively achieving precise left-right swaying and forward-backward pitching movements. This independent division of labor design ensures the precise controllability of the two posture movements, enabling the output of single-direction movements independently while also providing a foundation for subsequent complex movements. This breaks through the limitation of existing technologies that can only achieve single movements, enriching the robot's posture expression capabilities.

[0018] The rotational motion module is specifically designed to drive the robot to rotate around its central axis, adding a new dimension to the overall rotational motion. This design enables the robot to not only achieve localized posture changes but also complete overall orientation adjustments. Combined with swaying and pitching movements, it can create multi-dimensional composite effects, enhancing the dynamic hierarchy and visual appeal of the motion, and meeting the needs of various scenarios for the robot's all-around dynamic performance.

[0019] The main controller is connected to three motion modules, supporting independent or collaborative drive modes and enabling flexible switching between motion modes. When driven independently, the robot can perform basic movements such as pure swaying, pure pitching, or rotation in place, adapting to simple interactive scenarios. When driven collaboratively, it can combine various complex movements such as swaying and pitching, and swaying / pitching and rotation, adapting to display scenarios requiring strong visual impact and greatly improving the system's scene adaptability.

[0020] Through independent and collaborative control of each module, the robot can perform a variety of multi-degree-of-freedom compound movements, completely solving the problem of the single movement form in existing technologies. The diverse compound movements make the robot's dynamic performance richer and more interesting, effectively attracting crowd attention, creating a lively interactive atmosphere, and fully meeting the core needs of commercial marketing, exhibitions, entertainment and other scenarios for robots to be highly attractive and generate strong buzz.

[0021] This application achieves diversified outputs of multi-degree-of-freedom composite motion through an integrated layout, independently functional motion modules, and flexible collaborative control. The various components work together to ensure both precise control of individual motions and smooth coordination of composite motions, significantly enriching the robot's motion performance and effectively enhancing its attractiveness and interactivity in target scenarios. The design is simple, highly practical, and possesses outstanding application value.

[0022] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This paper shows a schematic diagram of the structure of a multi-degree-of-freedom motion robot system provided in Embodiment 1 of this application; Figure 2 This paper shows a disassembled schematic diagram of a multi-degree-of-freedom motion robot system structure provided in Embodiment 2 of this application; Figure 3 This paper illustrates the hardware interaction architecture of a multi-degree-of-freedom motion robot system provided in Embodiment 2 of this application; Figure 4 A flowchart of a control method for a multi-degree-of-freedom motion robot system provided in Embodiment 2 of this application is shown; Figure 5 A flowchart of a cooperative motion trajectory generation method provided in Embodiment 2 of this application is shown; Figure 6 A flowchart of a control instruction generation method provided in Embodiment 2 of this application is shown; Figure 7 A flowchart of a synchronization coordination processing method provided in Embodiment 2 of this application is shown. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0026] Example 1 To facilitate understanding of this application, the following is combined with... Figure 1 The flowchart of the multi-degree-of-freedom motion robot system provided in Embodiment 1 of this application illustrates Embodiment 1 in detail.

[0027] See Figure 1 As shown, Figure 1 A schematic diagram of a multi-degree-of-freedom motion robot system provided in Embodiment 1 of this application is shown. The multi-degree-of-freedom motion robot system includes a robot upper body, a connecting mechanism, and a mobile chassis. The mobile chassis integrates a left-right swaying motion module, a forward-backward swaying motion module, a rotational motion module, and a main controller.

[0028] Specifically, the robot system is applied in commercial marketing, display, and entertainment scenarios. The three motion execution modules integrated into the mobile chassis are all independently designed and work together to generate compound motions, namely the left and right swaying motion module, the forward and backward swaying motion module, and the rotation motion module.

[0029] The main controller, or main control board, is the core of the entire system. It contains two core components: an edge computing processor and a motion control unit. The former runs on a Linux operating system, while the latter runs on a FreeRTOS real-time operating system. Together, they achieve perception processing, instruction generation, and motion driving functions.

[0030] The mobile chassis is also equipped with a battery to power the entire system, and integrates auxiliary drive components such as stepper motor drivers and DC brushless motor drivers to ensure the stable operation of each motion module.

[0031] The left-right swaying motion module and the forward-backward swaying motion module are connected to the upper body of the robot through the connecting mechanism, and are used to drive the upper body of the robot to achieve left-right swaying motion and forward-backward pitching motion, respectively.

[0032] Specifically, the connecting mechanism has a built-in ball screw. The first slide driver and the second slide driver are connected to the corresponding ball screws through couplings. The motor transmits power through the couplings, driving the ball screws to move the slides, ensuring accurate power transmission and smooth movement.

[0033] The left and right swaying motion module consists of left and right slides and corresponding slide drivers. The left and right slides are installed in the horizontal plane of the robot body, and the sliding direction is consistent with the left and right direction of the robot. The slide drivers control the robot to make reciprocating linear motion along the axis, which drives the upper body of the robot to sway left and right around the central axis, simulating the human "head shaking" or "left and right swaying body" movements.

[0034] The forward and backward rocking motion module consists of front and rear slides and corresponding slide drivers. The front and rear slides are also installed in the horizontal plane of the robot body. The sliding direction is consistent with the forward and backward direction of the robot and perpendicular to the axis of the left and right slides. The slide drivers control the robot to make reciprocating linear motion along the axis, causing the upper body of the robot to pitch back and forth around the central axis, simulating the human "nodding" or "pitching back and forth" movements.

[0035] The rotational motion module is used to drive the robot to rotate around its central axis.

[0036] Specifically, the rotary motion module consists of a hub motor and a hub driver connected to it. The hub motor is mounted on the robot chassis to form a rotating platform. Its rotating part is an independent platform above the chassis. The cross slider structure composed of the left and right slides and the front and rear slides is mounted on this independent platform.

[0037] The independent platform rotates by relying on the output shaft of the hub motor. The hub motor is controlled by the hub driver to drive the robot to rotate clockwise or counterclockwise around the central axis, providing basic rotational functions for compound motion.

[0038] The main controller is connected to the left and right swaying motion module, the forward and backward swaying motion module, and the rotation motion module respectively. It is used to drive the three motion modules independently or in concert according to control commands, so that the robot can perform multi-degree-of-freedom compound motions including pure swaying, pure pitching, swaying and pitching combined, in-place rotation, and swaying / pitching and rotation combined.

[0039] Specifically, the main controller is connected to the slide driver (which controls the left and right, front and rear slides) and the wheel hub driver (which controls the wheel hub motor) via communication lines. The control commands come from preset programs, sensor signals (such as detection signals from human millimeter-wave radar sensors, human infrared sensors, etc.) or host computer commands.

[0040] The main controller generates coordinated control signals that can drive the three motion modules to work independently, achieving single movements such as pure left-right swaying, pure forward-backward pitching, and stationary rotation; it can also drive the modules to work together to achieve complex movements such as swaying and pitching, or swaying / pitching and rotation, for example, swaying left and right while rotating clockwise to produce a dynamic effect of "joyful dancing".

[0041] This multi-modal motion design solves the problem of monotonous motion in existing marketing robots. By combining anthropomorphic movements with composite motions, it enhances visual appeal and interactivity.

[0042] In one optional implementation, the left-right swaying motion module includes left and right slides and a first slide driver that drives them; the front-back swaying motion module includes front and back slides and a second slide driver that drives them; and the rotational motion module includes a hub motor and a hub driver that drives it.

[0043] Specifically, both the first and second slide drivers are slide drivers. The left and right slides and the front and rear slides are linear modules. They adopt a cross slider structure design with upper and lower layers, and are installed in the center based on the center point of the robot.

[0044] The limiting design ensures that the sliding direction of the left and right slides is consistent with the left and right direction of the robot, and the sliding direction of the front and rear slides is consistent with the front and rear direction of the robot, thus avoiding movement deviation and ensuring precise action.

[0045] As the power source for rotational motion, the hub motor, in conjunction with the hub drive, can precisely control the rotational speed and direction, providing stable power support for the overall rotational motion.

[0046] The left and right slides and the front and rear slides are cross-slider structures stacked vertically, and are installed centered on the center point of the robot.

[0047] Specifically, the cross slider structure belongs to the motion execution module of the mobile chassis. The linear module design makes its motion control precise and its response speed fast. Moreover, the structure is mature and compact, which can be easily integrated into the robot body without the need for complex exposed linkage mechanisms, thus ensuring the robot's clean appearance.

[0048] The centered installation method ensures the stability of the robot's center of gravity during movement, avoids movement imbalance caused by eccentricity, and provides a symmetrical motion basis for compound movements in the left-right and forward-backward directions, thereby improving motion coordination.

[0049] The connecting mechanism includes a ball screw, and the first slide driver and the second slide driver are respectively connected to the corresponding ball screws via couplings.

[0050] Specifically, the coupling connects the output shaft of the drive unit to the ball screw, effectively transmitting power, reducing losses and deviations during power transmission, and ensuring the accuracy of the slide table's movement.

[0051] As a precision transmission component, the ball screw features high transmission efficiency and high positioning accuracy. When combined with the control of the slide table driver, it can achieve the smoothness and precision of the reciprocating linear motion of the slide table, ensuring the realism of anthropomorphic movements.

[0052] The rotating part of the rotating motion module is an independent platform located above the chassis. The cross slider structure is installed on this independent platform, and the independent platform rotates by means of the output shaft of the hub motor.

[0053] Specifically, the independent platform is the core load-bearing component for the rotational motion. It not only supports the cross-slider structure but also drives the robot's upper body to rotate together. Its direct connection with the output shaft of the hub motor reduces intermediate transmission links and improves the response speed and control accuracy of the rotational motion.

[0054] This structural design allows rotational motion to be independent of, yet coordinated with, swaying and pitching motions, providing a structural foundation for the realization of compound motion and ensuring flexible switching between multiple motion modes.

[0055] In an optional implementation, the main controller includes an edge computing processor for processing perceived data and interaction information, identifying user intent and scene context, and generating high-level motion intent commands containing emotional semantics and motion patterns.

[0056] Specifically, the edge computing processor is equipped with a Linux operating system, and its core functions include camera video stream acquisition, network connection (supporting 4G / WIFI), immersive interaction with AI voice large model, facial recognition (which can recognize age, gender, emotions, including anger, disgust, aversion, fear, apprehension, happiness, joy, sadness, surprise, normal, etc.), and interaction with motion control unit.

[0057] The edge computing processor is connected to an interactive perception subsystem, which provides multimodal input for the generation of high-level motion intention commands and performs output feedback. The subsystem consists of a front-facing face recognition camera, a rear-facing face recognition camera, a directional microphone array, a display, dual-channel speakers, touch sensors, etc.

[0058] Edge computing processors dynamically generate or adjust high-level motion intent commands by fusing and analyzing visual information captured by cameras, audio information captured by microphone arrays, and touch signals captured by touch sensors. At the same time, they coordinate the output of motion control units with displays and speakers to achieve synchronous linkage of motion, vision, and audio.

[0059] The motion control unit is communicatively connected to the edge computing processor. It is used to receive the high-level motion intention command, calculate it into an executable coordinated motion trajectory of the left and right slides, front and rear slides and the rotating platform, and run a synchronous control algorithm and real-time closed-loop control to drive the first slide driver, the second slide driver and the hub motor.

[0060] Specifically, the motion control unit is equipped with the FreeRtos real-time operating system, which supports multi-tasking concurrent functions. Its core responsibilities include acquiring various sensor signals, implementing dual-slide control algorithms and turntable control algorithms, and receiving and processing various functional instructions issued by the edge processor.

[0061] The motion control unit is connected to a motion sensing subsystem, which provides high-precision feedback for real-time closed-loop control. The subsystem consists of a nine-axis attitude sensor, two laser trackers, a temperature and humidity sensor, a human millimeter-wave radar sensor, and a human infrared sensor.

[0062] The dual-slide control algorithm includes a master-slave control strategy and cross-coupling control (CCC). In master-slave control, one slide is the master slide that moves according to the position command, while the slave slide tracks the position of the master slide and compensates for synchronization errors. Cross-coupling control calculates the synchronization error of the position errors of the two slides and generates compensation amounts to correct the position commands or control outputs of the two slides respectively, ensuring synchronous movement.

[0063] The core formula of cross-coupling control includes the synchronization error calculation formula: es=e1-e2, where "es" represents the synchronization error of the two slides, "e1" represents the position error of slide 1 (the difference between the commanded position 1 and the actual position 1), and "e2" represents the position error of slide 2 (the difference between the commanded position 2 and the actual position 2).

[0064] The calculation formula for the compensation amount of cross-coupling control is: u_sync1=-K_sync*es, u_sync2=K_sync*es, where "u_sync1" represents the synchronization compensation amount of slide 1, "u_sync2" represents the synchronization compensation amount of slide 2, and "K_sync" represents the synchronization control gain (which can be dynamically adjusted according to the motion scenario).

[0065] The final control output formulas for cross-coupling control are: u1 = u1_pid + u_sync1, u2 = u2_pid + u_sync2, where "u1" represents the final control output of slide 1, "u2" represents the final control output of slide 2, "u1_pid" represents the output of the position loop PID controller of slide 1, and "u2_pid" represents the output of the position loop PID controller of slide 2.

[0066] If the correction method of "adding the synchronization compensation amount to the position command" is adopted, the core formula is: u_sync=K_sync*es, ref1=ref-u_sync, ref2=ref+u_sync, where "u_sync" represents the unified synchronization compensation amount, "ref1" represents the corrected position command of slide 1, "ref2" represents the corrected position command of slide 2, and "ref" represents the original target position command of the two slides (ref is the same when moving synchronously).

[0067] In an alternative implementation, the edge computing processor is connected to an interaction-aware subsystem for providing multimodal inputs for generating the high-level motion intent commands and for performing output feedback.

[0068] Specifically, the interactive perception subsystem's cameras are divided into front-facing and rear-facing face recognition cameras, which are responsible for collecting visual information such as user facial features and movement trajectories to provide data support for face recognition and scene judgment; the directional microphone array is mainly responsible for sound pickup, which can recognize audio signals from multiple directions and accurately pick up user voice commands.

[0069] The display is mainly used for multimedia presentation and human-computer interaction interface presentation. It can also display corresponding facial expressions based on the user's emotions recognized by the edge computing processor. The dual-channel speakers are mainly responsible for audio playback, including voice output for AI voice interaction, background music playback, etc., providing audio feedback for interaction.

[0070] Touch sensors can detect user touch operations and trigger sound, light, electricity, facial expressions, and voice displays as feedback, enriching the ways of human-computer interaction and enhancing the interactive experience.

[0071] The interactive perception subsystem includes: a camera for acquiring visual information; a microphone array for acquiring audio information; and a display and speakers for outputting visual and audio feedback.

[0072] Specifically, the visual information captured by the camera is transmitted to the edge computing processor, which combines it with AI algorithms for facial recognition to accurately determine the user's age, gender, and emotional state, providing emotional semantic basis for the generation of high-level motion intention commands; the audio information captured by the directional microphone array is processed to realize voice command recognition, enabling the robot to respond to the user's voice interaction needs.

[0073] The display's facial expression display function is based on visual information captured by dual cameras and algorithms trained on a large model. It can dynamically adjust facial expressions according to the user's emotions, such as displaying a happy expression when the user is happy. The audio output of the speaker works in conjunction with the motion control unit's motion control, such as playing cheerful music during "happy dance" movements to enhance the immersive experience of the interaction.

[0074] The edge computing processor is used to fuse and analyze the visual information and the audio information, dynamically generate or adjust the high-level motion intention commands, and coordinate the output of the motion control unit with the display and the speaker.

[0075] Specifically, edge computing processors can accurately identify user intentions and scene context by fusing multimodal information such as vision, audio, and touch. For example, when a user touches a robot and expresses curiosity in voice, it generates a high-level motion intention command of "leaning slightly forward" and coordinates the display to show the curious expression and the speaker to output a response voice.

[0076] When the scene changes, the edge computing processor can dynamically adjust the instructions. For example, when the user changes from happy to sad, the motion mode can be adjusted from "rapid small-amplitude swaying" to slow pitching, the display expression can be switched synchronously, and the speaker volume and tone can be adjusted accordingly to achieve adaptive interaction.

[0077] In an optional implementation, the motion control unit is connected to a motion sensing subsystem for providing high-precision feedback for the real-time closed-loop control.

[0078] Specifically, the nine-axis attitude sensor of the motion perception subsystem can determine the robot's current attitude in real time, including roll angle, pitch angle, and yaw angle. The motion control unit uses this attitude data to coordinate and control the stepper motor in a closed loop to achieve various precise swaying movements.

[0079] Two laser trackers monitor the position of the dual slides in real time, mainly for calibration and compensation calculations. They are used to verify the slide position data collected by the displacement sensors to ensure the accuracy of position feedback. Temperature and humidity sensors collect the temperature and humidity data of the current environment and report them to the central processing unit in real time. The data can be provided to the user through an LCD screen or voice dialogue.

[0080] The human-sensing millimeter-wave radar sensor and human infrared sensor are mainly responsible for detecting the human body and tracking its trajectory. The underlying motion control unit processes the acquired multi-target data and reports it to the central processing unit, providing a scenario basis for the robot's motion adjustment, such as adjusting the rotation direction according to the human body trajectory to face the user.

[0081] The motion sensing subsystem includes: an attitude sensor for monitoring the real-time attitude of the robot body; and a displacement sensor for monitoring the real-time position of the left and right slides and the front and rear slides.

[0082] Specifically, the attitude sensor is a nine-axis attitude sensor, which can collect attitude data to provide real-time feedback on the robot's motion status. When the robot's attitude deviates, the motion control unit can adjust the control commands in a timely manner to ensure accurate movements. The displacement sensor collects the actual positions of the left and right slides and the front and rear slides in real time and compares them with the commanded positions to provide data support for position error calculation and compensation.

[0083] The monitoring data of the laser tracker on the position of the dual slides can be used to correct the measurement error of the displacement sensor, improve the accuracy of position feedback, provide reliable input data for the synchronous control algorithm and error compensation module, and ensure the accuracy and smoothness of the composite motion.

[0084] The motion control unit is used to perform dynamic compensation and adjustment during the calculation and execution of the cooperative motion trajectory based on the attitude and position feedback data, so as to ensure the accuracy and smoothness of the composite motion.

[0085] Specifically, dynamic compensation and adjustment are achieved through the collaborative implementation of multiple algorithms, including enhanced PID control algorithm, collaborative control module, and error compensation module. The enhanced PID control algorithm includes fuzzy adaptive PID and anti-saturation processing mechanism.

[0086] The core calculation formula for enhanced PID control is: output = pid->kp*err + pid->ki*pid->integral + pid->kd*pid->derivative, where "output" represents the final output of the PID controller, "pid->kp" represents the proportional coefficient, "pid->ki" represents the integral coefficient, "pid->kd" represents the derivative coefficient, "err" represents the current position error (the difference between the commanded position and the actual position), "pid->integral" represents the integral term (accumulated error value), and "pid->derivative" represents the derivative term (related value of the error rate of change).

[0087] The formula for calculating the derivative filter of the enhanced PID is: d=(err-pid->prev_err) / pid->dt, pid->derivative=pid->derivative*(1-pid->d_filter)+d*pid->d_filter, where "d" represents the original derivative calculation value, "pid->prev_err" represents the position error of the previous control cycle, "pid->dt" represents the control cycle (fixed operation time interval), and "pid->d_filter" represents the derivative filter coefficient (used to suppress noise interference).

[0088] The collaborative control module adopts a virtual coupling model and a dynamic error allocation algorithm. The core formula for dynamic error allocation is: total_err=fabs(err_x)+fabs(err_y), importance_x=fabs(err_x), importance_y=fabs(err_y), max_allocation=total_err*0.8, where "total_err" represents the total error between the X and Y axes, "err_x" represents the position error of the X axis (such as the left and right slides), "err_y" represents the position error of the Y axis (such as the front and rear slides), "importance_x" represents the X-axis error sensitivity, "importance_y" represents the Y-axis error sensitivity (here, the absolute value of the error is used as the basic sensitivity calculation), and "max_allocation" represents the maximum total amount of allocable error (keeping a 20% margin).

[0089] The final allocation formula for dynamic error allocation is: When importance_x > importance_y * 1.5, alloc_x=min_double(max_allocation*0.7,fabs(err_x))、 alloc_y=min_double(max_allocation*0.3,fabs(err_y)); When importance_y > importance_x * 1.5, alloc_x=min_double(max_allocation*0.3,fabs(err_x))、 alloc_y=min_double(max_allocation*0.7,fabs(err_y)); Otherwise, alloc_x = min_double(max_allocation * 0.5, fabs(err_x)). alloc_y=min_double(max_allocation*0.5,fabs(err_y)), Here, "alloc_x" represents the X-axis allocation error, "alloc_y" represents the Y-axis allocation error, and "min_double(a,b)" represents an auxiliary function that returns the smaller of a and b.

[0090] The direction correction formula for dynamic error allocation is: \*redist_x=(err_x>0)?alloc_x:-alloc_x, \*redist_y=(err_y>0)?alloc_y:-alloc_y, where "\*redist_x" represents the final corrected X-axis error allocation value (including direction), and "\*redist_y" represents the final corrected Y-axis error allocation value (including direction).

[0091] The error compensation module employs a nonlinear compensation network and a 4D compensation grid query algorithm, with bilinear interpolation as the core calculation method. Its basic formulas include: dx=x1-x0, dy=y1-y0, wx1=(x1-x) / dx, wx0=(x-x0) / dx, wy1=(y1-y) / dy, wy0=(y-y0) / dy, where "dx" represents the grid side length in the X direction, "dy" represents the grid side length in the Y direction, "x0" and "y0" represent the coordinates of the lower left corner of the grid, "x1" and "y1" represent the coordinates of the upper right corner of the grid, "x" and "y" represent the coordinates of the target point, "wx1" and "wx0" represent the interpolation weights in the X direction, and "wy1" and "wy0" represent the interpolation weights in the Y direction.

[0092] The formula for calculating the compensation value of bilinear interpolation is: x_comp0=wx1*q00.x_comp+wx0*q10.x_comp、 x_comp1=wx1*q01.x_comp+wx0*q11.x_comp、 result.x_comp=wy1*x_comp0+wy0*x_comp1、 y_comp0=wx1*q00.y_comp+wx0*q10.y_comp、 y_comp1=wx1*q01.y_comp+wx0*q11.y_comp、 result.y_comp=wy1*y_comp0+wy0*y_comp1, Where “q00.x_comp” and “q00.y_comp” represent the X / Y direction compensation values ​​of the bottom left corner point (x0, y0) of the grid, “q01.x_comp” and “q01.y_comp” represent the X / Y direction compensation values ​​of the top left corner point (x0, y1) of the grid, “q10.x_comp” and “q10.y_comp” represent the X / Y direction compensation values ​​of the bottom right corner point (x1, y0) of the grid, “q11.x_comp” and “q11.y_comp” represent the X / Y direction compensation values ​​of the top right corner point (x1, y1) of the grid, “x_comp0” and “x_comp1” represent the intermediate interpolation results in the X direction, “y_comp0” and “y_comp1” represent the intermediate interpolation results in the Y direction, and “result.x_comp” and “result.y_comp” represent the final X / Y direction compensation values ​​of the target point.

[0093] See Figure 2 As shown, Figure 2 This document illustrates a disassembled schematic diagram of a multi-degree-of-freedom motion robot system structure provided in Embodiment 2 of this application. Its core value lies in transforming the abstract technical architecture into an intuitive and practical hardware visualization. It not only clearly presents the overall hardware layout of the multi-degree-of-freedom motion robot, but also accurately demonstrates the assembly relationship of core components such as the slide driver, main control board, left and right slides, front and rear slides, battery, wheel hub driver, and wheel hub motor on the mobile chassis, intuitively and completely reflecting the structural composition of the system.

[0094] This disassembly diagram achieves two key implementations: First, it visualizes the three-layer architecture of "robot body, connecting mechanism, and mobile chassis" from Example 1 in hardware; second, it presents the four core modules—"left-right swaying motion module, forward-backward swaying motion module, rotational motion module, and main controller"—at the hardware level, constructing a three-level visual association of "architecture-module-component." The mobile chassis, as the core hardware carrier of the entire system, overcomes the industry pain points of traditional robot hardware being scattered and having complex wiring. Through highly integrated design, it not only integrates the core execution components of the three motion modules, including left and right slides, forward and backward slides, and hub motors, but also integrates the drive components, including slide drivers and hub drivers, the control core (main control board), and the power supply unit (battery), achieving full-link integrated execution, drive, control, and power supply, significantly improving the system's integration and reliability.

[0095] Meanwhile, the disassembly diagram clearly defines the hardware composition logic of each motion module. The left-right swaying motion module consists of left and right slides and corresponding slide drivers; the forward-backward swaying motion module consists of forward and backward slides and corresponding slide drivers; and the rotational motion module consists of hub motors and connected hub drivers. It also clearly demonstrates the physical assembly relationship between the main controller (main control board) and each drive component, successfully transforming the abstract module definition into a hardware operational relationship that can directly guide production assembly. Furthermore, the disassembly view allows for a direct observation that the left and right slides and the forward and backward slides adopt a stacked cross-slider structure design, the hub motors are directly connected to the independent platform, and all core motion components are centrally positioned based on the center point of the moving chassis. This design ensures the stability of the robot's center of gravity during movement while maximizing the use of chassis space, effectively overcoming the technical bottlenecks of excessive hardware size and motion imbalance in multi-degree-of-freedom motion robots, and providing solid hardware structural support for the system to achieve precise and stable multi-degree-of-freedom motion.

[0096] See Figure 3 As shown, Figure 3 This paper illustrates the hardware interaction architecture of a multi-degree-of-freedom motion robot system provided in Embodiment 2 of this application. Its core value lies in clearly presenting the complete connection relationship and collaborative logic of each component of the system. It deconstructs the hardware interaction relationship between the dual-core design of the main controller ("edge computing processor combined with motion control unit") and the interaction perception subsystem and motion perception subsystem, providing intuitive hardware interaction guidance for the implementation of the technical solution.

[0097] The architecture diagram clearly shows the connection paths of all core components: the edge computing processor is directly connected to the display, human infrared sensor, WIFI, 4G, dual-channel speaker, rear face recognition camera, front face recognition camera and MIC array; the motion control unit is connected to the stepper motor driver, DC brushless motor driver, expression management board, human millimeter-wave radar sensor, temperature and humidity sensor, nine-axis attitude sensor, touch sensor and two laser trackers. The stepper motor driver is responsible for driving the X-axis slide and Y-axis slide, the DC brushless motor driver is responsible for driving the turntable, and the expression management board is specifically for controlling the breathing light. Through such precise connections, the interaction logic between the main controller and the interactive perception and motion perception subsystems is clearly defined.

[0098] In terms of functional division, the edge computing processor undertakes the core responsibility of "perception-decision". It realizes remote communication through WIFI and 4G modules and connects to the backend server and language big model link. At the same time, it collects multimodal interaction data, obtains visual information with the help of front and rear face recognition cameras, identifies multi-directional audio signals by picking up sound through directional MIC array, detects the presence of human bodies in the environment by human infrared sensor, and captures touch operations by touch sensor. Then, it presents multimedia content and interactive interface through display, plays audio by relying on dual-channel speakers, and controls breathing light to display facial expressions by combining expression management board, forming a complete interactive closed loop of "collection-processing-feedback".

[0099] The motion control unit serves as the core of the "execution-control" process. It is equipped with the FreeRTOS real-time operating system to support multi-task concurrency. It connects to actuators such as the X-axis slide, Y-axis slide, and turntable through stepper motor drivers and brushless DC motor drivers to precisely drive various motion actions. At the same time, it acquires rich motion feedback data, realizes human detection and trajectory tracking through a human-sensor millimeter-wave radar sensor, collects ambient temperature and humidity data with the help of temperature and humidity sensors, uses a nine-axis attitude sensor to determine the robot's roll angle, pitch angle, yaw angle, and other attitudes in real time, and relies on two laser trackers to monitor the position of the dual slides in real time for calibration and compensation calculations to achieve high-precision motion control.

[0100] The clear division of labor and collaboration between the two cores effectively overcomes the response delay problem caused by the mixed "perception-control" functions of traditional single controllers. The architecture diagram also fully presents the hardware composition of the two major subsystems: the "input-output" hardware combination of the interactive perception subsystem includes dual cameras for vision, a microphone array and dual-channel speakers for audio, touch sensors for touch, and human infrared sensors for environmental perception; the "feedback-calibration" hardware combination of the motion perception subsystem includes a nine-axis attitude sensor for attitude monitoring, two laser trackers for position monitoring, a temperature and humidity sensor for environmental data acquisition, and a human-tracking millimeter-wave radar sensor. This provides a clear hardware carrier for the technical solutions of "multimodal input to generate high-level motion intent" and "high-precision feedback to achieve closed-loop control" in Example 1, successfully solving the technical pain point of the disconnect between perception and control.

[0101] Meanwhile, the architecture diagram clarifies the communication paths of all execution components. The X-axis slide and Y-axis slide are connected to the motion control unit through stepper motor drivers. The turntable communicates with the motion control unit through a DC brushless motor driver. The breathing light is controlled by the motion control unit through the expression management board. All execution components are connected to the control core through dedicated drive modules, effectively avoiding signal interference. This provides a reliable hardware interaction foundation for the stable implementation of synchronous control algorithms and dynamic compensation algorithms, and overcomes the technical difficulties of signal chaos and poor synchronization when multiple components work together.

[0102] Example 2 Based on the same application concept, see [link / reference] Figure 4 As shown, Figure 4 A flowchart of a control method for a multi-degree-of-freedom motion robot system provided in Embodiment 2 of this application is shown. The method, applied to the multi-degree-of-freedom motion robot system as shown in Embodiment 1 above, includes steps S401-S405: S401: Receive target motion command.

[0103] The sources of target motion commands are diverse. They can be fixed motion pattern commands stored in preset programs (such as "dancing happily" or "nodding to greet"), custom commands issued by the host computer, or dynamic commands generated by the edge computing processor based on multimodal information (user voice commands, touch operations, facial emotions, etc.) collected by the interactive perception subsystem.

[0104] The multimodal information acquisition of the interactive perception subsystem relies on a front-facing face recognition camera, a rear-facing face recognition camera (to acquire visual information of facial emotions), a directional microphone array (to acquire audio information of voice commands), and a touch sensor (to acquire touch operation signals). The edge computing processor is equipped with a Linux operating system and uses an AI voice model and a face recognition algorithm (which can recognize age, gender, and emotions such as anger, disgust, aversion, fear, apprehension, happiness, joy, sadness, surprise, and normal) to convert multimodal information into dynamic commands.

[0105] During the instruction reception process, the edge computing processor and the motion control unit work together. The edge computing processor is responsible for receiving external multimodal information and converting it into instructions, while the motion control unit is responsible for receiving preset programs or host computer instructions, ensuring the comprehensiveness and real-time nature of instruction reception.

[0106] S402: Based on the target motion command, generate a coordinated motion trajectory for the left-right swaying motion module, the forward-backward swaying motion module, and the rotation motion module.

[0107] The generation process needs to take into account multiple factors, including the emotional semantics of the high-level motion intention commands, the robot's current posture state (feedback from the nine-axis posture sensor), the real-time motion state of each motion module (slide position, rotation angle, etc.), and kinematic constraints.

[0108] The nine-axis attitude sensor provides real-time feedback on the robot's roll angle, pitch angle, and yaw angle. The position of the slide is collected by a combination of two laser trackers and displacement sensors, and the rotation angle is collected by the hub motor encoder. The emotional semantics of the high-level motion intention commands need to be matched with the corresponding motion parameters. For example, the emotion of "happy" corresponds to "rapid small-amplitude swaying + rotation", and the emotion of "calm" corresponds to "slow forward and backward pitch + rotation".

[0109] Kinematic constraints include the maximum travel limit of the slide, the maximum speed limit of the motor, the jerk limit, and the range of rotation angles, ensuring that the generated cooperative motion trajectory is within the hardware's operating range and avoiding equipment damage or malfunction due to exceeding hardware limits. The generation of the cooperative motion trajectory requires ensuring that the three motion modules coordinate with each other in time and space; for example, the speed and phase of the swinging and rotating motions must be coordinated to avoid motion conflicts and achieve a smooth composite motion effect.

[0110] S403: Perform smooth planning on the cooperative motion trajectory and generate control commands for each motion module.

[0111] The core of the smooth planning algorithm is the S-shaped velocity curve algorithm, which is based on the continuous function of the third derivative jerk. It achieves smooth start and stop through a seven-segment trajectory planning. The seven segments are: acceleration stage (jerk+, increase acceleration), uniform acceleration stage (a=max, maintain maximum acceleration), deceleration stage (jerk-, decrease acceleration to 0), uniform speed stage (v=max, maintain maximum speed), acceleration-deceleration stage (jerk-, increase deceleration), uniform deceleration stage (a=-max, maintain maximum deceleration), and deceleration stage (jerk+, decrease deceleration to 0).

[0112] By constraining the jerk during motion, motion shocks caused by sudden acceleration can be avoided, allowing the speed to smoothly increase from 0 to its maximum value and then smoothly decrease to 0, while the position changes continuously over time, forming a smooth motion profile without abrupt changes or impacts, thus improving the fluidity and anthropomorphism of the movement.

[0113] Based on the smooth motion profile and combined with the hardware parameters of each motion module (such as the transmission ratio of the slide table, the rated speed of the motor, etc.), the target position, target speed and target acceleration of each driver (first slide table driver, second slide table driver, hub driver) in each control cycle are calculated. These parameters are the control commands of each motion module, which can directly drive the hardware to execute motion.

[0114] The control cycle is a fixed operation cycle of the motion control unit. The slide transmission ratio and the rated speed of the motor are known hardware parameters. Through trajectory planning of the S-shaped speed curve, the target parameters in each cycle can be accurately calculated to ensure that the command and the smooth contour are highly matched.

[0115] The smoothing planning process also requires the use of a forward control algorithm, the core of which is path curvature calculation and look-ahead point selection. The curvature calculation formula is: curvature = θ / L, where "θ" is the angle difference between adjacent path points (θ = arccos((dx1×dx2+dy1×dy2) / (norm1×norm2)), "dx1" and "dy1" are the x and y direction differences of the previous path segment, "dx2" and "dy2" are the x and y direction differences of the next path segment, "norm1" and "norm2" are the vector magnitudes of the corresponding path segments, and "L" is the average value of the adjacent path segments. The average length is (L=(norm1+norm2) / 2); the look-ahead point is determined by the relaxation factor relaxation=0.8-0.6×clamped_curvature; where "relaxation" represents the relaxation factor for the look-ahead point selection, "clamped_curvature=min(max_curvature / 0.5,1.0)", "max_curvature" represents the maximum curvature of the path, and "clamped_curvature" is the curvature value after clamping.

[0116] S404: Synchronize and coordinate the control commands used to drive the left-right swaying motion module, the forward-backward swaying motion module, and the rotation motion module.

[0117] Specifically, the core of the synchronous coordination processing targets the control commands of the left-right swaying motion module and the forward-backward swaying motion module. It adopts a master-slave control strategy or cross-coupled control (CCC) algorithm to ensure the synchronization of the two slides' movements, while also taking into account the motion coordination with the rotational motion module to avoid conflicts between multiple modules.

[0118] The core logic of the master-slave control strategy is as follows: a master slide moves according to position commands, and a slave slide tracks the position of the master slide and corrects errors through synchronization compensation; the core formula of the cross-coupling control (CCC) algorithm is: Position error: e1=ref1-pos1, e2=ref2-pos2; where “e1” represents the position tracking error of slide 1, “ref1” represents the target position command of slide 1, and “pos1” represents the actual position of slide 1; “e2” represents the position tracking error of slide 2, “ref2” represents the target position command of slide 2, and “pos2” represents the actual position of slide 2.

[0119] Synchronization error: es = e1 - e2, where "es" represents the synchronization error between the two slides, "e1" represents the position error of slide 1 (the difference between the commanded position 1 and the actual position 1), and "e2" represents the position error of slide 2 (the difference between the commanded position 2 and the actual position 2).

[0120] Synchronization compensation amount: u_sync1=-K_sync*es, u_sync2=K_sync*es, where "u_sync1" represents the synchronization compensation amount of slide 1, "u_sync2" represents the synchronization compensation amount of slide 2, and "K_sync" represents the synchronization control gain (which can be dynamically adjusted according to the motion scenario).

[0121] The final control output formulas are: u1 = u1_pid + u_sync1, u2 = u2_pid + u_sync2, where "u1" represents the final control output of slide 1, "u2" represents the final control output of slide 2, "u1_pid" represents the output of the position loop PID controller of slide 1, and "u2_pid" represents the output of the position loop PID controller of slide 2.

[0122] The formula for correcting the position command by adding the synchronization compensation amount is: u_sync=K_sync*es, ref1=ref-u_sync, ref2=ref+u_sync, where "u_sync" represents the unified synchronization compensation amount, "ref1" represents the corrected position command of slide 1, "ref2" represents the corrected position command of slide 2, and "ref" represents the original target position command of the two slides (ref is the same when moving synchronously).

[0123] The synchronous coordination process relies on real-time data fed back by the motion sensing subsystem, including the real-time position of the dual slides (acquired by displacement sensors and laser trackers) and the real-time angle of the rotating platform (acquired by the hub motor encoder). Closed-loop synchronous control is achieved through data feedback.

[0124] For the control commands of the rotary motion module, the synchronous coordination process mainly adjusts its rotational speed and phase to match the rhythm of the swaying motion. For example, when the swaying motion is fast, the rotational speed increases accordingly, and when the swaying motion is slow, the rotational speed decreases, thus achieving coordination of the compound motion.

[0125] S405: Output control commands and perform real-time closed-loop adjustments based on sensor feedback to execute the multi-degree-of-freedom composite motion.

[0126] Specifically, the output control commands are transmitted to each driver via communication lines. After receiving the target position, speed, and acceleration parameters, the first and second slide drivers drive the slides to perform reciprocating linear motion. After receiving the control commands, the hub driver drives the hub motor to rotate, which in turn drives the rotating platform to rotate. All modules work together to perform compound motion.

[0127] The instruction format received by the driver is "target position + target speed + target acceleration". The stage driver drives the ball screw through a coupling to move the stage, and the hub driver controls the rotation speed and direction of the hub motor to achieve smooth rotation of the rotating platform.

[0128] The sensor feedback data comes from multiple sensors of the motion perception subsystem, including the attitude data of the nine-axis attitude sensor, the stage position data of the laser tracker and displacement sensor, the environmental data of the temperature and humidity sensor, the human trajectory data of the human presence sensor, etc. These data are transmitted to the motion control unit in real time to provide a basis for closed-loop adjustment.

[0129] There are two core methods for real-time closed-loop adjustment: one is the closed-loop servo control that enhances the PID control algorithm, and the other is the feedforward compensation based on the multi-dimensional compensation model. The two methods work together to ensure motion accuracy and smoothness.

[0130] The core formula and mechanism of the enhanced PID control algorithm are as follows: PID output calculation: output = Kp * err + Ki * ∫err × dt + Kd * (Δerr / Δt), where "output" represents the final output of the PID controller (driver control signal), "Kp" represents the proportional coefficient, "Ki" represents the integral coefficient, "Kd" represents the differential coefficient, "err" represents the current position error (err = ref - pos), "∫err × dt" represents the integral term (error accumulation value), and "Δerr / Δt" represents the differential term (error change rate).

[0131] Fuzzy adaptive PID parameter correction formula: Kp' = Kp + ΔKp, Ki' = Ki + ΔKi, Kd' = Kd + ΔKd. Here, "Kp'", "Ki'", and "Kd'" represent the corrected PID parameters, "Kp", "Ki", and "Kd" represent the original PID parameters, and "ΔKp", "ΔKi", and "ΔKd" represent the parameter correction amounts output by the fuzzy logic controller (the inputs are the position error err and the error change rate ec = Δerr / Δt).

[0132] Integral anti-windup backstepping formula: When output > output_max and err > 0, ∫err × dt = ∫err × dt - err × dt; when output < output_min and err < 0, ∫err × dt = ∫err × dt - err × dt. Here, "output_max" represents the upper limit of the PID output, "output_min" represents the lower limit of the PID output, "∫err × dt" represents the integral accumulation value, "err" represents the current position error, and "dt" represents the control period.

[0133] Differential filtering calculation formula: der=(1-α)*der_prev+α*(Δerr / Δt) where “der” represents the differential term after filtering, “α” represents the filtering coefficient, “der_prev” represents the differential term of the previous period, and “Δerr / Δt” represents the current rate of change of the original error.

[0134] The core of feedforward compensation based on the multi-dimensional compensation model is 4D compensation mesh query and bilinear interpolation, and its formula and logic are as follows: 4D compensation mesh dimensions: spatial position ((x,y), corresponding to the slide table coordinates), ambient temperature ((temp), motor load (load); Temperature zone interpolation formula: temp_factor=(temp-T_low) / (T_high-T_low), comp_temp=comp_temp0×(1-temp_frac)+comp_temp1×temp_frac where "temp_factor" represents the temperature normalization factor (clamped to [0,1]), "temp" represents the current ambient temperature, "T_low" represents the minimum temperature threshold, "T_high" represents the maximum temperature threshold, "comp_temp" represents the compensation amount after temperature interpolation, "comp_temp0" represents the compensation amount of the current temperature zone, "comp_temp1" represents the compensation amount of the next temperature zone, and "temp_frac=temp_factor×2-temp_idx" (temp_idx is the temperature zone index 0~2).

[0135] The formulas for calculating load partition weights are: weight_i = exp(-0.5 × ((load - load_i) / σ_load)²) and weight_normi = weight_i / ∑(i = 0~3)weight_i, where “weight_i” represents the weight of the i-th load partition, “load” represents the current motor load, “load_i” represents the preset i-th load level value (4 in total), “σ_load” represents the load standard deviation, and “weight_normi” represents the normalized load weight.

[0136] Bilinear interpolation (spatial location) formula: Interpolation in the X direction: comp_x0 = comp_x0y0 × (x1 - x) / (x1 - x0) + comp_x1y0 × (x - x0) / (x1 - x0), comp_x1 = comp_x0y1 × (x1 - x) / (x1 - x0) + comp_x1y1 × (x - x0) / (x1 - x0) Final interpolation in the Y direction: comp_x = comp_x0 × (y1 - y) / (y1 - y0) + comp_x1 × (y - y0) / (y1 - y0) Where “comp_x0y0”, “comp_x1y0”, “comp_x0y1”, and “comp_x1y1” represent the X-direction compensation amounts of the four corner points (x0,y0), (x1,y0), (x0,y1), and (x1,y1) of the grid, respectively. “x0” and “y0” represent the coordinates of the lower left corner point of the grid, “x1” and “y1” represent the coordinates of the upper right corner point of the grid, “x” and “y” represent the coordinates of the target point, “comp_x0” and “comp_x1” represent the intermediate interpolation results in the X direction, and “comp_x” represents the final X-direction compensation amount of the target point (the calculation logic for the Y-direction compensation amount is the same; the formula only needs to replace x with y).

[0137] Total feedforward compensation formulas: comp_totalx = comp_tempx + comp_loadx + comp_x, comp_totally = comp_tempy + comp_loady + comp_y Where “comp_totalx” and “comp_totaly” represent the total compensation amount of the X and Y axes, respectively; “comp_tempx” and “comp_tempy” represent the temperature interpolation compensation amount; “comp_loadx” and “comp_loady” represent the load interpolation compensation amount; and “comp_x” and “comp_y” represent the spatial position bilinear interpolation compensation amount.

[0138] This control method is fully compatible with the robot system hardware architecture of Example 1. Steps S401-S405 cover the entire process from instruction reception to motion execution, fully combining the structural characteristics of the hardware and the control logic of the algorithm to ensure the coordinated implementation of hardware functions and control strategies, achieving precise execution of multi-degree-of-freedom composite motion. It solves the problems of single-function motion control, poor synchronization, and unstable movements in existing robots. Through steps such as trajectory planning, synchronization coordination, and closed-loop adjustment, it improves the accuracy and flexibility of motion control, adapting to the needs of various scenarios such as commercial marketing, demonstration, and entertainment.

[0139] In an optional implementation, see Figure 5 As shown, Figure 5 The flowchart illustrates a method for generating a coordinated motion trajectory according to Embodiment 2 of this application. The step of generating a coordinated motion trajectory for the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module based on the target motion command includes steps S501-S502: S501: Parse the target motion command and generate a high-level motion intention command that includes motion patterns and emotional semantics.

[0140] Specifically, the instruction parsing adopts the logic of "keyword extraction + emotion mapping". Keyword extraction identifies the motion type (such as "swing", "rotation", "compound") from the target instruction. Emotion mapping matches the motion parameters with the emotion recognition results of the interactive perception subsystem to ensure that the emotion expression is consistent with the motion characteristics.

[0141] The mapping relationship between emotional semantics and motion parameters is determined through large model training. For example, "anger" corresponds to "large-amplitude rapid swaying + slow rotation", and "surprise" corresponds to "small-amplitude rapid pitching + paused rotation", so that the robot's actions can convey the corresponding emotional color.

[0142] S502: Based on the high-level motion intention command, the current robot state, and kinematic constraints, calculate the trajectory data for the spatiotemporal coordination of the left-right swaying motion module, the forward-backward swaying motion module, and the rotation motion module, and use this data as the coordinated motion trajectory.

[0143] Specifically, the current robot state includes the initial position of the slide, the initial rotation angle, the initial load of the motor, etc. These parameters need to be substituted as the initial conditions for trajectory calculation; kinematic constraints need to be transformed into mathematical boundaries to ensure that the trajectory parameters are within the range allowed by the hardware.

[0144] The trajectory data calculation needs to ensure that the parameters of the three motion modules are coordinated in time and space. For example, the motion phases of the left and right slides and the front and rear slides are coordinated, and the rotation angle and the swing amplitude are matched to form a smooth composite motion trajectory.

[0145] In an optional implementation, see Figure 6 As shown, Figure 6 The flowchart of a control command generation method provided in Embodiment 2 of this application is shown, wherein the step of smoothing the cooperative motion trajectory and generating control commands for each motion module includes steps S601 to S603: S601: Using the S-shaped velocity curve algorithm, the trajectory segments corresponding to the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module in the cooperative motion trajectory are planned respectively.

[0146] Specifically, trajectory segment planning needs to be divided into seven stages, and the acceleration, jerk, and velocity parameters of each stage are set according to the S-curve algorithm to ensure smooth transition between each trajectory segment.

[0147] The trajectory segment planning logic of the rotation module is the same as that of the front and rear slides. Only the position parameters need to be replaced with rotation angles and the speed with angular velocities to ensure that the trajectory planning standards of all motion modules are consistent.

[0148] S602: By constraining the jerk during the motion process, a smooth motion profile with continuously changing acceleration, velocity and position is generated.

[0149] Specifically, the jerk constraint is achieved through a seven-stage jerk threshold, ensuring that the acceleration curve is continuous without abrupt changes, the velocity curve is a smooth S-shape, and the position curve is a smooth parabola without any jerking.

[0150] S603: Based on the smooth motion profile, calculate the target position, target speed, and target acceleration of the first slide driver of the left-right swaying motion module, the second slide driver of the front-back swaying motion module, and the hub driver of the rotational motion module in each control cycle, and use them as control commands for each motion module.

[0151] Specifically, the control cycle is a fixed operation cycle of the motion control unit. The target parameters for each cycle are obtained through relevant calculations of the smooth motion contour, ensuring that the instructions can directly drive the hardware to execute and are highly consistent with the planned trajectory.

[0152] In an optional implementation, see Figure 7 As shown, Figure 7 The flowchart of a synchronization coordination processing method provided in Embodiment 2 of this application is shown, wherein the synchronization coordination processing of the control commands used to drive the left-right swaying motion module and the forward-backward swaying motion module includes steps S701 to S705: S701: Obtain the real-time motion status of the left-right swaying motion module and the forward-backward swaying motion module.

[0153] Specifically, the real-time motion status is acquired through a dual-source system of "displacement sensor + laser tracker". The acquisition frequency is consistent with the control cycle, and the acquired parameters include the actual position and speed of the slide, ensuring that the status data is comprehensive and reliable.

[0154] Outlier removal is required when collecting data. If the difference between the collected data and the previous period is too large, it is considered an anomaly, and the data from the previous period is used to avoid abnormal data affecting synchronous control.

[0155] S702: Calculate the tracking error between the motion state of the two modules and their corresponding control commands.

[0156] Specifically, tracking error includes position error and velocity error, with position error being the core. The calculation formula is as follows: Left and right slide position tracking error: e1=ref1-pos1, where “e1” represents the position tracking error of slide 1, “ref1” represents the target position command of slide 1, and “pos1” represents the actual position of slide 1.

[0157] Front and rear slide position tracking error: e2=ref2-pos2, where "e2" represents the position tracking error of slide 2, "ref2" represents the target position command of slide 2, and "pos2" represents the actual position of slide 2.

[0158] Position error must be controlled within a reasonable range; if it exceeds this range, synchronization compensation will be triggered.

[0159] S703: Calculate the synchronization error between the tracking errors of the two modules.

[0160] Specifically, the synchronization error is calculated based on the position tracking error, with the core formula: es=e1-e2, where "es" is the synchronization error, reflecting the relative deviation between the two slides; the allowable threshold for the synchronization error needs to be set according to the synchronization accuracy, and if it is exceeded, a synchronization compensation amount needs to be generated.

[0161] S704: Generate synchronization compensation amount based on the synchronization error.

[0162] Specifically, the synchronization compensation amount is generated by a proportional (P) controller, with the core formula: u_sync = K_sync × e_s, where "u_sync" is the synchronization compensation amount, "K_sync" is the synchronization control gain, and "e_s" is the synchronization error between the two slides.

[0163] If cross-coupling control is used, bidirectional compensation quantities need to be generated: u_sync1 = -K_sync × e_s, u_sync2 = K_sync × e_s, where "u_sync1" is the compensation quantity for the left and right slides, "u_sync2" is the compensation quantity for the front and rear slides, "K_sync" is the synchronization control gain, and "e_s" is the synchronization error between the two slides. The compensation quantities need to be limited in amplitude to avoid over-compensation leading to motion oscillation.

[0164] S705: The control commands for the left-right swaying motion module and the forward-backward swaying motion module are corrected using the synchronization compensation amount to suppress the synchronization error of the two modules during the motion process.

[0165] Specifically, the instruction correction is for the target position instruction, and the correction formula is: the instruction after correction for the left and right slides: ref_1'=ref_1+u_sync1, where "ref_1'" is the corrected target position, "ref_1" is the original target position instruction for the left and right slides, and "u_sync1" is the compensation amount for the left and right slides.

[0166] The corrected command for the front and rear slides is: ref_2' = ref_2 + u_sync2, where "ref_2'" is the corrected target position, "ref_2" is the original target position command for the front and rear slides, and "u_sync2" is the compensation amount for the front and rear slides. The corrected command needs to be sent to the driver in the next control cycle to achieve real-time closed-loop control and ensure that the synchronization error remains stable within a reasonable range.

[0167] In an optional implementation, the real-time closed-loop adjustment based on sensor feedback to execute the multi-degree-of-freedom composite motion includes: employing an enhanced PID control algorithm for closed-loop servo control, the algorithm including integral anti-saturation and differential filtering mechanisms.

[0168] Specifically, the implementation of enhanced PID control algorithm relies on the real-time computing power of motion control unit. Its core steps and formulas are as follows: Error calculation: err=ref'-pos_cal, where "err" is the current position error, "ref'" is the corrected target position command (including synchronization compensation and feedforward compensation), and "pos_cal" is the calibrated actual position (average value of displacement sensor + laser tracker data).

[0169] Integral term calculation and anti-saturation: integral = integral + err × T, where "integral" is the integral cumulative value, "err" is the current position error, and "T" is the control period; Integral limit: if (integral > I_max) integral = I_max; else if (integral < -I_max) integral = -I_max, where "integral" is the integral cumulative value, "I_max" is the upper limit of integral, and "-I_max" is the lower limit of integral; If output > output_max and err > 0, then integral = integral - err × T (integral rollback), where "integral" is the integral cumulative value, "err" is the current position error, "T" is the control period, and "output_max" is the upper limit of PID output.

[0170] Differential term calculation and filtering: der_raw=(err-err_prev) / T (original differential), where “der_raw” is the original differential term, “err” is the current position error, “err_prev” is the error of the previous cycle, and “T” is the control cycle; der=der×(1-α)+der_raw×α (filtered differential), where “der” is the filtered differential term (the left side is the filtered differential term of the previous cycle), “α” is the filtering coefficient, and “der_raw” is the current original differential term.

[0171] PID output: output = Kp × err + Ki × integral + Kd × der, where “output” represents the final output of the PID controller (driver control signal), “Kp” represents the proportional coefficient, “Ki” represents the integral coefficient, “Kd” represents the derivative coefficient, “err” represents the current position error, “integral” represents the integral term (accumulated error value), and “der” represents the derivative term (error change rate after filtering).

[0172] Fuzzy adaptive adjustment: Based on the magnitude of the error and the rate of change of the error, the PID parameters are dynamically adjusted through fuzzy logic to ensure a fast response when the error is large and a stable, oscillation-free response when the error is small.

[0173] Alternatively, feedforward compensation based on a multi-dimensional compensation model can be used to offset errors caused by system nonlinearity and environmental disturbances.

[0174] Specifically, the core of the multi-dimensional compensation model is the 4D compensation mesh and bilinear interpolation, and its complete implementation logic and formula are as follows: 4D Compensation Mesh Construction: Error data under different working conditions are collected through offline calibration experiments, and the error data is stored in the 4D mesh as compensation.

[0175] Real-time operating condition matching: During the movement, the current spatial position, ambient temperature, and motor load are collected in real time, and the grid index and weight are calculated.

[0176] Multi-dimensional interpolation calculation of compensation amount: The total compensation amount is calculated through temperature interpolation, load interpolation, and bilinear interpolation (spatial location).

[0177] Feedforward compensation application: The total compensation is superimposed on the target position command to offset the errors caused by system nonlinearity and environmental disturbances, thereby improving motion accuracy.

[0178] The multi-degree-of-freedom motion robot system provided in this application embodiment can be specific hardware on a device or software or firmware installed on the device. The system provided in this application embodiment has the same implementation principle and technical effects as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the system embodiment section can be referred to the corresponding content in the aforementioned method embodiment. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0179] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some communication interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0180] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0181] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0182] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0183] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A multi-degree-of-freedom motion robot system, characterized in that, The multi-degree-of-freedom motion robot system includes a robot upper body, a connecting mechanism, and a mobile chassis; the mobile chassis integrates a left-right swing motion module, a forward-backward swing motion module, a rotation motion module, and a main controller. The left-right swaying motion module and the forward-backward swaying motion module are connected to the upper body of the robot through the connecting mechanism, and are used to drive the upper body of the robot to achieve left-right swaying motion and forward-backward pitching motion respectively. The rotational motion module is used to drive the robot as a whole to rotate around its central axis; The main controller is connected to the left and right swaying motion module, the forward and backward swaying motion module, and the rotation motion module respectively. It is used to drive the three motion modules independently or in concert according to control commands, so that the robot can perform multi-degree-of-freedom compound motions including pure swaying, pure pitching, swaying and pitching combined, in-place rotation, and swaying / pitching and rotation combined.

2. The system according to claim 1, characterized in that, The left and right swaying motion module includes left and right slides and a first slide driver that drives them; the front and back swaying motion module includes front and back slides and a second slide driver that drives them; and the rotation motion module includes a hub motor and a hub driver that drives it. The left and right slides and the front and rear slides are cross-slider structures stacked vertically, and are installed centered based on the center point of the robot. The connecting mechanism includes a ball screw, and the first slide driver and the second slide driver are respectively connected to the corresponding ball screws through couplings; The rotating part of the rotating motion module is an independent platform located above the chassis. The cross slider structure is installed on this independent platform, and the independent platform rotates by means of the output shaft of the hub motor.

3. The system according to claim 2, characterized in that, The main controller includes: Edge computing processors are used to process perceived data and interaction information, identify user intent and scene context, and generate high-level motion intent commands containing emotional semantics and motion patterns. The motion control unit is communicatively connected to the edge computing processor. It is used to receive the high-level motion intention command, calculate it into an executable coordinated motion trajectory of the left and right slides, front and rear slides and the rotating platform, and run a synchronous control algorithm and real-time closed-loop control to drive the first slide driver, the second slide driver and the hub motor.

4. The system according to claim 3, characterized in that, The edge computing processor is connected to an interactive perception subsystem, which provides multimodal input for generating the high-level motion intention command and performs output feedback. The interactive sensing subsystem includes: a camera for acquiring visual information; a microphone array for acquiring audio information; and a display and speakers for outputting visual and audio feedback. The edge computing processor is used to fuse and analyze the visual information and the audio information, dynamically generate or adjust the high-level motion intention commands, and coordinate the output of the motion control unit with the display and the speaker.

5. The system according to claim 3, characterized in that, The motion control unit is connected to a motion sensing subsystem, which provides high-precision feedback for the real-time closed-loop control. The motion sensing subsystem includes: an attitude sensor for monitoring the real-time attitude of the robot body; and a displacement sensor for monitoring the real-time positions of the left and right slides and the front and rear slides. The motion control unit is used to perform dynamic compensation and adjustment during the calculation and execution of the cooperative motion trajectory based on the attitude and position feedback data, so as to ensure the accuracy and smoothness of the composite motion.

6. A control method for a multi-degree-of-freedom motion robot system, characterized in that, The method, applied to a multi-degree-of-freedom motion robot system as described in any one of claims 1-5, comprises: Receive target motion instructions; Based on the target motion command, generate a coordinated motion trajectory for the left-right swaying motion module, the forward-backward swaying motion module, and the rotation motion module; The coordinated motion trajectory is smoothly planned to generate control commands for each motion module; The control commands used to drive the left and right swaying motion module, the forward and backward swaying motion module, and the rotation motion module are processed synchronously and in a coordinated manner. Output control commands and perform real-time closed-loop adjustments based on sensor feedback to execute the multi-degree-of-freedom composite motion.

7. The method according to claim 6, characterized in that, The step of generating a coordinated motion trajectory for the left-right swaying motion module, the forward-backward swaying motion module, and the rotation motion module based on the target motion command includes: The target motion command is parsed to generate a high-level motion intent command that includes motion patterns and emotional semantics; Based on the high-level motion intention command, the current robot state, and kinematic constraints, trajectory data for the spatiotemporal coordination of the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module is calculated and used as the coordinated motion trajectory.

8. The method according to claim 6, characterized in that, The smoothing of the coordinated motion trajectory and the generation of control commands for each motion module include: The S-shaped velocity curve algorithm is used to plan the trajectory segments corresponding to the left-right swaying motion module, the forward-backward swaying motion module, and the rotational motion module in the cooperative motion trajectory. By constraining the jerk during the motion process, a smooth motion profile with continuously changing acceleration, velocity, and position is generated. Based on the smooth motion profile, the target position, target velocity, and target acceleration of the first slide driver of the left-right swaying motion module, the second slide driver of the front-back swaying motion module, and the hub driver of the rotational motion module are calculated in each control cycle, and used as control commands for each motion module.

9. The method according to claim 6, characterized in that, The process of synchronizing and coordinating the control commands used to drive the left-right swaying motion module and the forward-backward swaying motion module includes: Obtain the real-time motion status of the left-right swaying motion module and the forward-backward swaying motion module; Calculate the tracking error between the motion state of the two modules and their corresponding control commands; Calculate the synchronization error between the tracking errors of the two modules; A synchronization compensation amount is generated based on the synchronization error; The control commands for the left-right swaying motion module and the forward-backward swaying motion module are corrected using the synchronization compensation amount to suppress the synchronization error between the two modules during the motion process.

10. The method according to claim 6, characterized in that, The real-time closed-loop adjustment based on sensor feedback to execute the multi-degree-of-freedom composite motion includes: An enhanced PID control algorithm is used for closed-loop servo control, and the algorithm includes integral anti-saturation and differential filtering mechanisms. Alternatively, feedforward compensation based on a multi-dimensional compensation model can be used to offset errors caused by system nonlinearity and environmental disturbances.

Images