Massage robot, control method, and massage method

Abstract

A massage robot and a control method therefor. The massage robot comprises a robot body (1). The robot body (1) comprises an information acquisition portion (4), a system control portion (2), and a massage execution portion (3). The information acquisition portion (4) acquires an image of a region to be massaged. The system control portion (2) plans a massage trajectory, a pose trajectory, and a massage force level. The massage execution portion (3) performs massage on the basis of the plan set by the system control portion (2). By fitting the massage trajectory to match the contours of the human body, smooth transitions between respective curve segments are achieved, and throughout the entire motion process, the intermediate segments in the trajectory are all traversed at a constant speed, achieving smooth and stable movement, and maintaining constant-force pressing to achieve the massage effect.

Description

Massage robot, control method and massage method

[0001] Cross-references

[0002] This application is based on the Chinese patent application with application number 202311033198.9 and application date August 16, 2023, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is hereby introduced into this application as a reference. Technical Field

[0003] The present invention relates to the technical field of robot massage, and in particular to a massage robot, a control method and a massage method. Background Art

[0004] Massage involves applying specialized techniques to specific areas or acupuncture points to induce physiological responses through physical action, unblocking meridians and balancing body functions. Massage can boost metabolism and is a widely used health care method. Popular massage devices on the market today are massage chairs and massage beds. However, both are passive devices, controlled from the bottom up. The user relies on their own body weight to massage, requiring them to adjust their muscles to "seek" the pressure points. The massaged area is under the pressure of their own body weight and cannot relax. Furthermore, traditional massage chairs and massage beds offer limited techniques and cannot simulate and teach massage movements, often failing to achieve the desired results.

[0005] Intelligent robotics technology has rapidly developed over the years. It offers interactive functions, with the key goal of integrating into society and serving people's lives. It also boasts tireless and immense power. Based on Traditional Chinese Medicine massage techniques, incorporating artificial intelligence, intelligent sensing, robotics, and modern rehabilitation theory, massage robots have emerged. These robots can assist or even replace masseuses in performing conventional massage techniques. Their precise force / position feedback control allows for more standardized massage techniques and more consistent results. Leveraging intelligent robotics, research into massage robots is a key tool in modernizing massage. Therefore, integrating modern intelligent robotics with traditional massage theory is a key development direction for massage robots.

[0006] As early as 1996, Masao Kume et al. in Japan designed a mechanical unit (MTU) with four degrees of freedom. Using a dual-loop control strategy, it performed shoulder massage movements, demonstrating the feasibility of developing an intelligent massage robot. From 2009 to 2010, researchers at National Taiwan University in my country proposed an impedance control method for a multi-finger manipulator based on myoelectric analysis. By recording the electromyogram (EMG) of the forearm during massage and determining the relationship between human sensation and massage force, they determined the appropriate input command range for the manipulator during massage. Ultimately, they achieved effective and comfortable massage using a multi-finger manipulator. However, these technologies relied on manual input data and lacked adaptive adjustments, resulting in a low level of intelligence.

[0007] Summary of the Invention

[0008] (1) Purpose of the invention

[0009] The purpose of the present invention is to provide a massage robot, control method and massage method that can improve the degree of intelligence, do not require manual positioning, ensure constant force pressing in the direction of contact between the massage execution part and the human body, and improve the massage effect.

[0010] (2) Technical solution

[0011] To solve the above problems, the present invention provides a massage robot, a robot body, the robot body including an information acquisition unit, a system control unit and a massage execution unit; the information acquisition unit, the system control unit and the massage execution unit are electrically connected;

[0012] The information acquisition unit acquires an image of the area to be massaged;

[0013] The system control unit plans a massage trajectory, a posture trajectory and a massage force position according to the image acquired by the information acquisition unit; and the massage execution unit executes the massage according to the plan of the system control unit.

[0014] In another aspect of the present invention, preferably, the system control unit includes a model training subsystem, a trajectory generation subsystem, a posture control subsystem, and a force-position hybrid control subsystem;

[0015] The model training subsystem obtains three-dimensional point cloud data of the general massage area through deep learning training, the trajectory generation subsystem plans the massage trajectory according to the image of the area to be massaged collected by the information acquisition unit, the posture control subsystem generates a posture trajectory according to the planned massage trajectory, and the force-position hybrid control subsystem plans the massage force position according to the posture trajectory.

[0016] In another aspect of the present invention, preferably, the massage execution part includes a massage mechanical arm, a massage head and a heating component; the massage head is arranged at the end of the massage mechanical arm, the heating component is arranged inside the massage head, and the heating component is used to heat the massage head.

[0017] In another aspect of the present invention, preferably, a control method of the massage robot as described above includes:

[0018] acquiring an image of the area to be massaged;

[0019] The obtained image of the area to be massaged is matched one-to-one with the three-dimensional point cloud data of the preset universal massage area to obtain the massage key points and postures on the image of the area to be massaged; the three-dimensional point cloud data of the preset universal massage area is obtained through training;

[0020] Planning a massage trajectory using a preset NURBS curve according to the massage key points on the image of the area to be massaged;

[0021] Use the preset force-position hybrid control algorithm to adjust the massage trajectory in real time;

[0022] The massage is completed according to the massage trajectory and force-position hybrid control algorithm.

[0023] In another aspect of the present invention, preferably,

[0024] The image of the area to be massaged includes a two-dimensional image and a three-dimensional image;

[0025] The obtained image of the area to be massaged is matched one-to-one with the three-dimensional point cloud data of the preset universal massage area to obtain massage key points on the image of the area to be massaged, including:

[0026] generating a two-dimensional trajectory and two-dimensional trajectory key points according to the two-dimensional image of the image of the area to be massaged;

[0027] Converting the two-dimensional trajectory key points into three-dimensional key points in a three-dimensional image of the area to be massaged according to a preset first conversion rule;

[0028] The three-dimensional key points are matched one-to-one with the three-dimensional point cloud data of the preset universal massage area to obtain the massage key points on the image of the area to be massaged.

[0029] In another aspect of the present invention, preferably, the acquired image of the area to be massaged is matched one-to-one with the three-dimensional point cloud data of the preset universal massage area to obtain the posture on the image of the area to be massaged, including:

[0030] Obtain the position and normal vector corresponding to the three-dimensional point cloud data and the massage key points;

[0031] The position and normal vector are converted into a posture according to a preset second conversion rule.

[0032] In another aspect of the present invention, preferably, the planned massage trajectory also includes adjusting the posture trajectory in the massage trajectory in real time using a preset posture algorithm; the preset posture algorithm includes an S-type acceleration and deceleration or a sinusoidal acceleration and deceleration control algorithm, and the S-type acceleration and deceleration or sinusoidal acceleration and deceleration control algorithm includes a trajectory look-ahead algorithm.

[0033] In another aspect of the present invention, preferably, the preset force-position hybrid control algorithm includes:

[0034] obtaining a force to be applied to the massage area;

[0035] calculating the acceleration of the massage execution part according to the force to be applied to the massage area;

[0036] Calculating the end position of the massage execution part according to the force and acceleration to be applied to the massage area;

[0037] The motion angle of the massage execution part is calculated according to the end position and the inverse kinematic solution of the robot arm.

[0038] In another aspect of the present invention, preferably, the preset force-position hybrid control algorithm calculation formula includes:

[0039] Calculate the acceleration,

[0040] Calculate the end pose,

[0041] Where F is the force applied to the human body, M represents the inertia parameter of impedance control, B represents the damping parameter of impedance control, K represents the stiffness parameter of impedance control, Δt is the call cycle, and x represents the position of the massage actuator. Indicates the speed of the massage execution unit. Indicates the acceleration of the massage execution unit.

[0042] Another aspect of the present invention is preferably a massage method, which uses the control method as described above to control a massage robot to perform massage.

[0043] (3) Beneficial effects

[0044] The above technical solution of the present invention has the following beneficial technical effects:

[0045] The massage robot of the present invention is provided with a system control unit to plan the massage trajectory, posture trajectory and massage force position, so that the overall massage process is closer to manual massage, and the massage effect is improved. Among them, the planned massage trajectory is fitted to the massage trajectory to meet the effect of fitting the human body trajectory, and a smooth transition between each curve segment is achieved; the posture trajectory allows each motion trajectory to accelerate, uniformly speed, and decelerate smoothly. During the entire motion process, the trajectory of the middle segment passes at a uniform speed, making the motion trajectory smoother; the massage force position ensures that in the direction of contact between the massage head and the human body, constant force is maintained to achieve a massage effect, and solves the problem that the point cloud position and posture recognized by the visual recognition are not fitted or squeezed due to their inherent errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] FIG1 is a schematic diagram of the overall structure of a massage robot according to an embodiment of the present invention;

[0047] FIG2 is a schematic diagram of the overall flow of a control method according to an embodiment of the present invention;

[0048] Reference numerals:

[0049] 1: Robot body; 2: System control unit; 3: Massage execution unit; 4: Information acquisition unit; 3-1: Massage robotic arm; 3-2: Massage head. DETAILED DESCRIPTION

[0050] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present invention.

[0051] The accompanying drawings illustrate schematic diagrams of layer structures according to embodiments of the present invention. These figures are not drawn to scale; for clarity, some details are exaggerated and some details may be omitted. The shapes, relative sizes, and positional relationships of the various regions and layers shown in the figures are merely exemplary and may deviate in practice due to manufacturing tolerances or technical limitations. Those skilled in the art may design regions / layers with different shapes, sizes, and relative positions based on actual needs.

[0052] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0053] In the description of the present invention, it should be noted that the terms "first", "second" and "third" are only used for descriptive purposes and should not be understood as indicating or implying relative importance.

[0054] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0055] The present invention will be described in more detail below with reference to the accompanying drawings. In each of the accompanying drawings, identical elements are represented by similar reference numerals. For the sake of clarity, the various parts in the accompanying drawings are not drawn to scale.

[0056] Example 1

[0057] A massage robot, FIG1 shows a schematic diagram of the overall structure of a massage robot according to an embodiment of the present invention, as shown in FIG1 , comprising: a robot body 1, the robot body 1 comprising an information acquisition unit 4, a system control unit 2 and a massage execution unit 3; the information acquisition unit 4, the system control unit 2 and the massage execution unit 3 are electrically connected; the information acquisition unit 4 acquires an image of the area to be massaged; the system control unit 2 plans a massage trajectory, a posture trajectory and a massage force position according to the area to be massaged acquired by the information acquisition unit; the massage execution unit 3 performs massage according to the plan of the system control unit; the specific contents of the information acquisition unit 4, the system control unit 2 and the massage execution unit 3 are not limited here, and the information acquisition unit 4 can be a two-dimensional camera or a depth camera Machine, optionally, in this embodiment, the specific content of the information acquisition unit 4 includes a depth camera for acquiring images of the area to be massaged. Further, in this embodiment, the content acquired by the information acquisition unit 4 includes a two-dimensional image and a three-dimensional image of the area to be massaged, and the position of the area to be massaged in the three-dimensional image is obtained by matching the two-dimensional image with the three-dimensional image. The information acquisition unit 4 is used to acquire the image of the area to be massaged; the specific content of the system control unit 2 is not limited here. Optionally, in this embodiment, the system control unit 2 includes a trajectory generation subsystem, a posture control subsystem and a force-position hybrid control subsystem; the trajectory generation subsystem plans the massage trajectory according to the image of the area to be massaged acquired by the information acquisition unit, and the posture control subsystem plans the massage trajectory according to the planned massage trajectory Generate posture trajectory, the force-position hybrid control subsystem plans massage force position according to the posture trajectory; further, the trajectory generation subsystem fits the massage trajectory to meet the effect of fitting the human body trajectory and achieve smooth transition between each curve segment; the posture control subsystem allows each motion trajectory to accelerate, uniformly speed, and decelerate smoothly. During the entire motion process, the trajectory of the middle segment passes at a uniform speed, making the motion trajectory smoother; the force-position hybrid control subsystem ensures that constant force is maintained in the direction of contact between the massage head and the human body. Furthermore, the system control unit also includes a model training subsystem, which obtains general three-dimensional point cloud data through training based on deep learning technology, and trains images of each massage area based on deep learning technology, so as to distinguish the areas that need massage. The location is which part of the human body, and where the boundary point is, so as to tailor a massage trajectory for each user; the specific content of the training is not limited here, and it can be based on pre-set user attribute items and massage service plan categories; sample behavior data corresponding to the user attribute items of the training samples are collected, and the massage service plan for the customer to be massaged is determined based on the attribute information corresponding to the user attribute items of the customer to be massaged and the massage service plan identification model to achieve the massage effect; the massage execution unit includes a massage robot arm and a massage head; the massage head is arranged at the end of the massage robot arm; the system control unit controls the movement of the massage robot arm according to the posture trajectory, and the system control unit controls the movement of the massage head according to the planned massage force position;Furthermore, the massage execution unit further includes a heating component, which is disposed inside the massage head and is used to heat the massage head to enhance the massage effect. The specific number of joints of the robotic arm and the specific structure of the massage head are not limited herein.

[0058] The information acquisition unit is arranged on the side of the end of the massage robot arm, and the specific location of the information acquisition unit is not limited here; it can be electrically connected to the outside of the massage robot arm, or it can be arranged on the inside of the massage robot arm, and the shooting angle and position can be adjusted by controlling the movement of the robot arm. Optionally, in this embodiment, the depth camera can be a binocular structured light infrared camera. Furthermore, the binocular structured light infrared camera is used to collect data from different perspectives such as looking down and looking straight. Furthermore, the specific model of the selected binocular structured light infrared camera is RealSense D435i.

[0059] The massage robot of this embodiment is provided with a trajectory generation subsystem, a posture control subsystem and a force-position hybrid control subsystem, so that the overall massage process is closer to manual massage and the massage effect is improved. Among them, the trajectory generation subsystem fits the massage trajectory to meet the effect of fitting the human body trajectory and achieve a smooth transition between each curve segment; the posture control subsystem allows each motion trajectory to accelerate, maintain a constant speed, and decelerate smoothly. During the entire motion process, the trajectory of the middle segment passes at a constant speed, making the motion trajectory smoother; the force-position hybrid control subsystem ensures that constant force is maintained in the direction of contact between the massage head and the human body to achieve a massage effect. Only by adding a deep learning-based visual recognition system can we truly tailor the massage trajectory for each user, overcome the shortcomings of separate trajectory planning that can only be applied to the movement of position and posture on fixed objects, and solve the problem that the visually recognized point cloud position and posture are not fitted or squeezed due to their inherent errors.

[0060] Example 2

[0061] A control method for the massage robot as described above, FIG2 shows an overall flow chart of the control method according to an embodiment of the present invention, as shown in FIG2 , includes:

[0062] Obtain an image of the area to be massaged; the image of the area to be massaged is collected by controlling the information collection unit 4, and the specific content of the collection can be a two-dimensional camera or a depth camera. Optionally, in this embodiment, the specific content of the information collection unit 4 includes a depth camera for collecting the image of the area to be massaged. Further, in this embodiment, the content of the image of the area to be massaged includes a two-dimensional image and a three-dimensional image of the area to be massaged; optionally, in this embodiment, the position of the human body in the three-dimensional image is obtained by matching the two-dimensional image with the three-dimensional image; further, the depth camera is used to simultaneously obtain a two-dimensional digital image and a three-dimensional depth image containing the human body image of the user's massaged part, the two-dimensional digital image is used for identifying the massage key points of the area to be massaged, and the three-dimensional depth image is used for three-dimensional inverse modeling to determine the three-dimensional spatial position of the massage key points; a digital image segmentation method is used to eliminate background noise, extract the image of the area to be massaged in the two-dimensional digital image, remove redundant image information, narrow the target range, reduce the amount of data, and improve computing efficiency; the position of the image of the area to be massaged in the three-dimensional depth image is obtained by matching the two-dimensional digital image with the three-dimensional depth image, eliminate background noise in the three-dimensional depth image, reduce the amount of data of the three-dimensional image, and improve computing efficiency;

[0063] The obtained image of the area to be massaged is matched one-to-one with the three-dimensional point cloud data of the preset universal massage area to obtain massage key points and postures on the image of the area to be massaged; the three-dimensional point cloud data of the preset universal massage area is obtained through training; the specific content of the three-dimensional point cloud data of the preset universal massage area is not limited here, and can be massage acupoint information applicable to the shoulder and neck, or can be overall massage acupoint information. The corresponding purpose is to find the massage key points and postures of the current masseur. Optionally, in this embodiment, obtaining the massage key points on the image of the area to be massaged includes: generating a two-dimensional trajectory and two-dimensional trajectory key points based on the two-dimensional image of the image of the area to be massaged; and converting the two-dimensional trajectory key points into massage key points in the three-dimensional image of the image of the area to be massaged according to a preset first conversion rule; The specific content of the first conversion rule is not limited here. Optionally, in this embodiment, a two-dimensional trajectory and two-dimensional trajectory key points are generated using a parametric equation, which can be a trajectory similar to a spiral circle. The two-dimensional trajectory key points are converted into massage key points in the three-dimensional image of the image of the area to be massaged using a transformation matrix; the massage key points are matched one-to-one with the three-dimensional point cloud data of the preset universal massage area; the specific content of the three-dimensional point cloud data of the preset universal massage area is not limited here, and optionally, it can be trained data; the position and normal vector corresponding to the three-dimensional point cloud data and the massage key points are obtained; the position and normal vector are converted into a posture according to a preset second conversion rule. The specific content of the preset second conversion rule is not limited here. Optionally, in this embodiment, the preset second conversion rule is to add the position and normal vector to the preset X-axis coordinate, generate the corresponding direction vector, and then convert it into a posture expression represented by the position plus the quaternion;

[0064] A massage trajectory is planned using preset NURBS curves based on key massage points on the image of the massage area. NURBS curves are used to plan and fit the massage trajectory to the human body. NURBS curves are a universal type of spline curve. Any point, line, arc, or free-form curve in three-dimensional space can be uniformly represented by this curve expression, demonstrating its generality and versatility. NURBS curves are based on B-spline curves, with weights added to their control points to enhance the curve's modeling capabilities. These curves can solve the connection and smoothing issues between complex curves. Therefore, NURBS curves are often used in CNC machine tool curve fitting and are a common curve and surface expression form currently used in computer-aided design software systems. Their excellent mathematical properties are used to plan complex trajectory curves in robot motion. Furthermore, in this embodiment, a preset posture algorithm is also used to generate a posture trajectory in the massage trajectory for real-time adjustment. The specific content of the preset posture algorithm is not limited here. Optionally, in this embodiment, the preset posture algorithm includes an S-type acceleration and deceleration or a sinusoidal acceleration and deceleration control algorithm to ensure smooth acceleration, uniform speed, and deceleration of each motion trajectory. In order to frequently start and stop, the preset posture algorithm also includes a trajectory look-ahead algorithm to minimize the deceleration and acceleration between each trajectory. During the entire motion process, the trajectory of the middle segment passes at a uniform speed, making the motion trajectory smoother. S-type acceleration and deceleration: The traditional S-type acceleration and deceleration model is based on the common velocity curve model and introduces the physical concept of acceleration to measure the flexibility of the mechanical system. The introduction of acceleration can reduce the impact on the motor and robot system during operation. The S-type acceleration and deceleration model can be divided into seven phases: acceleration phase, uniform acceleration phase, deceleration phase, uniform speed phase, acceleration and deceleration phase, uniform deceleration phase, and deceleration phase. The sinusoidal acceleration and deceleration control algorithm divides the motion process into acceleration and deceleration segments. The principle is as follows: during the acceleration phase, the acceleration a changes sinusoidally over time, from 0 to a maximum value, and then back to 0. Throughout the acceleration phase, the acceleration value remains on the positive half of the y-axis, and the velocity also reaches its maximum value during this phase, which is the uniform speed phase speed cm. During the uniform speed phase, the robot runs at a constant speed vm until the deceleration phase, which is also the main phase of the robot's end effector operation. During the deceleration phase, the acceleration a changes sinusoidally over time, from 0 to a minimum value, and then back to 0. During the entire acceleration phase, the acceleration value is always on the negative half of the y-axis, and the speed vm will also decrease to 0 during the uniform speed phase. At this time, the robot end effector completes the preset trajectory and reaches the target position, and the operation ends. The normal acceleration constraint is introduced into the trajectory look-ahead algorithm to limit the maximum feed rate of arc processing. By integrating the motion vector relationship, system dynamics performance, trajectory segment length and other constraints, the feed rate of the connection point is optimized to achieve a smooth transition of continuous trajectory segments during robot movement and improve stability.

[0065] The preset force-position hybrid control algorithm is used to plan the massage force position, and the massage trajectory is adjusted in real time according to the massage force position; the specific content of the preset force-position hybrid control algorithm is not limited here. Optionally, in this embodiment, the force-position hybrid control is realized by using the admittance impedance control, thereby realizing the effect of the normal constant force pressing of the massage head. Furthermore, in this embodiment, the preset force-position hybrid control algorithm is to obtain the force to be applied to the human body; calculate the acceleration of the massage execution part according to the force to be applied to the human body; calculate the end position of the massage execution part according to the force and acceleration to be applied to the human body, and calculate the movement angle of the massage execution part according to the end position and the kinematic inversion of the robotic arm; introduce force-position hybrid control to ensure that the constant force pressing is maintained in the direction of contact between the massage head and the human body to achieve the massage effect, and solve the problem that the position and posture of the point cloud recognized by the visual recognition are not fitted or squeezed due to its inherent error. Furthermore, in this embodiment, the calculation formula of the preset force-position hybrid control algorithm includes:

[0066] Calculate the acceleration,

[0067] Calculate the end pose,

[0068] Where F is the force to be applied to the human body, M represents the inertia parameter of impedance control, B represents the damping parameter of impedance control, K represents the stiffness parameter of impedance control, Δt is the call cycle, and x represents the position of the massage actuator. Indicates the speed of the massage execution part. Indicates the acceleration of the massage execution unit.

[0069] The position and posture of the point cloud identified by visual recognition will have a certain error. When running completely according to the planned trajectory, even a difference of 1mm from the target position will cause misalignment or squeezing. Therefore, force-position hybrid control is introduced to ensure that constant pressure is maintained in the direction of contact between the massage head and the human body to achieve the desired massage effect.

[0070] Massage is performed based on the posture trajectory and massage force position. A deep learning-based visual recognition algorithm, an irregular trajectory planning algorithm, and a force-position hybrid control algorithm work together to achieve coordinated "eye," "hand," "touch," and "brain." This avoids the monotony of a manual massage, the stop-and-go trajectory, and the fluctuating force. The result is a more balanced, smoother, and more personalized massage than manual massage.

[0071] Example 3

[0072] A massage method, using the control method as described above to control a massage robot to perform massage.

[0073] It should be understood that the above-described specific embodiments of the present invention are merely illustrative or illustrative of the principles of the present invention and do not constitute limitations of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made without departing from the spirit and scope of the present invention should be included within the scope of protection of the present invention. In addition, the appended claims are intended to cover all variations and modifications that fall within the scope and metes and bounds of the appended claims, or equivalents thereof.

[0074] While the above description does not provide detailed technical details regarding patterning and etching of each layer, those skilled in the art will appreciate that various conventional methods can be used to form layers, regions, and the like in desired shapes. Furthermore, those skilled in the art may devise methods that differ from those described above to achieve the same structure.

[0075] The present invention has been described above with reference to the embodiments thereof. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present invention. The scope of the present invention is defined by the appended claims and their equivalents. Those skilled in the art may make various substitutions and modifications without departing from the scope of the present invention, and such substitutions and modifications are intended to fall within the scope of the present invention.

Claims

1. A massage robot, characterized in that: include: A robot body (1), the robot body (1) comprising an information acquisition unit (4), a system control unit (2) and a massage execution unit (3); the information acquisition unit (4), the system control unit (2) and the massage execution unit (3) are electrically connected; The information acquisition unit (4) acquires an image of the area to be massaged; The system control unit (2) plans a massage trajectory, a posture trajectory and a massage force position according to the image acquired by the information acquisition unit; and the massage execution unit (3) executes the massage according to the plan of the system control unit.

2. The massage robot according to claim 1, characterized in that: The system control unit (2) comprises a model training subsystem, a trajectory generation subsystem, a posture control subsystem and a force-position hybrid control subsystem; The model training subsystem obtains three-dimensional point cloud data of a general massage area through deep learning training; the trajectory generation subsystem plans a massage trajectory according to the image of the massage area collected by the information acquisition unit; the posture control subsystem generates a posture trajectory according to the planned massage trajectory; and the force-position hybrid control subsystem plans the massage force position according to the posture trajectory.

3. The massage robot according to claim 1, characterized in that: The massage execution part (3) comprises a massage mechanical arm (3-1), a massage head (3-2) and a heating component; the massage head (3-2) is arranged at the end of the massage mechanical arm (3-1), the heating component is arranged inside the massage head (3-2), and the heating component is used to heat the massage head (3-2).

4. A control method for a massage robot as claimed in any one of claims 1 to 3, characterized in that: include: acquiring an image of the area to be massaged; The obtained image of the area to be massaged is matched one by one with the three-dimensional point cloud data of the preset universal massage area, and the massage key points and postures on the image of the area to be massaged are obtained; the three-dimensional point cloud data of the preset universal massage area is obtained through training; Planning a massage trajectory using a preset NURBS curve according to the massage key points on the image of the area to be massaged; Use the preset force-position hybrid control algorithm to adjust the massage trajectory in real time; The massage is completed according to the massage trajectory and the force-position hybrid control algorithm.

5. The control method according to claim 4, characterized in that: The image of the area to be massaged includes a two-dimensional image and a three-dimensional image; The obtained image of the area to be massaged is matched one by one with the three-dimensional point cloud data of the preset universal massage area, and massage key points on the image of the area to be massaged are obtained, including: Generate a two-dimensional trajectory and two-dimensional trajectory key points according to the two-dimensional image of the image of the area to be massaged; Converting the two-dimensional trajectory key points into three-dimensional key points in a three-dimensional image of the image of the area to be massaged according to a preset first conversion rule; The three-dimensional key points are matched one by one with the three-dimensional point cloud data of the preset universal massage area to obtain the massage key points on the image of the area to be massaged.

6. The control method according to claim 5, characterized in that: The obtained image of the area to be massaged is matched one by one with the three-dimensional point cloud data of the preset universal massage area to obtain the position and posture on the image of the area to be massaged, including: Obtain the positions and normal vectors corresponding to the three-dimensional point cloud data and the massage key points; The position and normal vector are converted into a posture according to a preset second conversion rule.

7. The control method according to claim 4, characterized in that: The planned massage trajectory also includes adjusting the posture trajectory in the massage trajectory in real time using a preset posture algorithm; the preset posture algorithm includes an S-type acceleration and deceleration or a sinusoidal acceleration and deceleration control algorithm, and the S-type acceleration and deceleration or sinusoidal acceleration and deceleration control algorithm includes a trajectory look-ahead algorithm.

8. The control method according to claim 4, characterized in that: The preset force-position hybrid control algorithm includes: obtaining a force to be applied to the massage area; calculating the acceleration of the massage execution part according to the force to be applied to the massage area; Calculating the end position of the massage execution part according to the force and acceleration to be applied to the massage area; The movement angle of the massage execution part is calculated according to the end position and the inverse solution of the robot arm kinematics.

9. The control method according to claim 8, characterized in that: The preset force-position hybrid control algorithm calculation formula includes: Calculate the acceleration, Calculate the end pose, Where F is the force applied to the human body, M represents the inertia parameter of impedance control, B represents the damping parameter of impedance control, K represents the stiffness parameter of impedance control, Δt is the call cycle, x represents the position of the massage actuator, Indicates the speed of the massage execution unit. Indicates the acceleration of the massage actuator.

10. A massage method, characterized in that: Use the control method described in any one of claims 4-9 to control the massage robot to massage.

Images