Massage robot intelligent control method and system

Abstract

The present application relates to the technical field of intelligent control, in particular to a massage robot intelligent control method and system, comprising the following steps: collecting contact pressure, spatial displacement and dynamic response information, extracting pressure gradient and depth change characteristics, identifying dynamic contact state and contact point interaction trend, and combining periodic feedback and deviation trend to continuously calibrate the basis and control parameters. The present application constructs a state criterion that is updated synchronously with the action period, places pressure change, spatial displacement and dynamic response in a unified analysis process, forms a continuous identification basis for contact evolution, and completes interaction trend recognition by combining the corresponding relationship between movement trend and force evolution, so that action adjustment can be orderly carried out in line with the current contact situation, and the continuous correction ability of the control basis is maintained by combining rhythm reorganization and deviation review, thereby improving force coordination, action transition smoothness and contact fit stability.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to an intelligent control method and system for a massage robot. Background Technology

[0002] Intelligent control technology involves real-time perception, analysis, and decision-making regarding the system's operating status, and autonomously adjusting equipment behavior based on specific control logic to achieve automatic control and management of complex objects or processes. It is widely applied in various fields such as industrial manufacturing, intelligent transportation, service robots, and medical rehabilitation and physiotherapy equipment. Among these, the traditional intelligent control method for massage robots refers to the way robots used in physiotherapy equipment control the massage process through preset programs or basic feedback mechanisms. This typically employs drive motors and pressure sensors, operating through fixed massage paths, frequencies, and intensities. Its control method relies on time settings, repetitive actions, or position sensor signals.

[0003] Existing technologies rely heavily on predetermined rhythms and repetitive trajectories to maintain massage output during operation. Their judgment of contact status is biased towards single signal feedback, making it difficult to continuously identify the linkage between force diffusion, contact point migration, and pressure depth. When faced with unevenness, softness, hardness, or changes in posture, the control basis is prone to being one-sided, resulting in a lack of targeted action transitions. Local areas are prone to imbalances in pressure, abrupt transitions, or slow compensation, which further leads to insufficient continuity of force application, unstable fit, fluctuations in comfort during the process, and restricts the ability to maintain stable control and make fine adjustments during the physiotherapy process. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and to propose an intelligent control method and system for massage robots.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent control method for a massage robot, comprising the following steps:

[0006] S1: Based on the end-effector massage component, analyze the contact data collected by the pressure sensor, compare the movement direction and pressure change trend, perform periodic difference analysis on the contact data, extract pressure gradient and depth change features, and obtain composite sensing state factors.

[0007] S2: Based on the composite sensing state factor, analyze the pressure gradient and continuous change characteristics of each contact area, compare the direction of motion with the direction of contact depth change, use the consistency of direction to determine the dynamic contact state, and obtain the contact interaction trend characteristics.

[0008] S3: Adjust the controller's response strategy to the touch interaction trend characteristics, combine the recognition state change execution logic, filter the control path that matches the current massage action, adjust the operation parameter allocation mode, and obtain the action mode conversion signal;

[0009] S4: Based on the action mode conversion signal, adjust the drive response relationship, update the action execution rhythm and ratio adjustment mode, switch to the new action state, and obtain a dynamic control instruction set;

[0010] S5: Based on the dynamic control instruction set, compare the contact feedback within the cycle, analyze the gradient change of the composite sensing state factor in the previous cycle, determine the current deviation trend, adjust the main control unit parameter settings and discrimination criteria, and obtain the synchronization deviation calibration result.

[0011] The present invention is improved in that the composite sensing state factor is a quantitative index obtained by weighted calculation based on signal synchronization, sensing accuracy and data fusion degree; the touch point interaction trend characteristics include action coordination, state recognition rate and interaction continuity; the action mode conversion signal includes switching stability, drive adaptability and parameter transmission consistency; the dynamic control instruction set includes action response rate, output synchronization and servo switching sensitivity; and the synchronization deviation calibration result includes deviation correction rate, calibration consistency and adjustment accuracy.

[0012] The present invention is improved in that the step of obtaining the composite sensing state factor is specifically as follows:

[0013] S111: Based on the end-effector massage component, analyze the contact data collected by the pressure sensor, monitor the spatial coordinate information output by the displacement detection unit, synchronously associate the pressure data with the spatial coordinates according to the time identifier, and construct the spatial position and pressure correspondence relationship by combining the actual distribution of the contact points to obtain the spatiotemporal pressure sequence of the contact points;

[0014] S112: Based on the spatiotemporal pressure sequence of the contact point, determine the motion direction vector of the end massage component, combine the pressure distribution change trend over a continuous period, compare the correspondence between the two types of direction vectors using the angle change method, associate the dynamic response information of the acceleration detection unit, and obtain the pressure direction consistency feature quantity.

[0015] S113: Based on the pressure direction consistency feature, the continuous periodic pressure distribution is processed by periodic differential processing, and the pressure gradient change between spatially adjacent contact points and the contact depth change range are combined to organize and aggregate multiple types of sensing information to obtain a composite sensing state factor.

[0016] The present invention is improved in that the step of obtaining the touch interaction trend features is specifically as follows:

[0017] S211: Based on the composite sensing state factor, analyze the pressure distribution of each contact point as a function of spatial coordinates, divide the region according to the spatial adjacency of the contact points, compare the direction of pressure change and the degree of continuity of distribution of adjacent contact points, and obtain the pressure gradient distribution characteristics.

[0018] S212: Based on the pressure gradient distribution characteristics, track the changing trend of each continuous spatial region on the time axis, calculate the changing direction of pressure distribution in each continuous region, compare the vector correspondence between the end motion direction and the contact depth change direction, and obtain the directional coordination relationship characteristics.

[0019] S213: Based on the directional coordination relationship characteristics, compare the synchronization between the pressure gradient distribution change and the contact depth change within the contact area, identify contact segments where the directional relationship remains continuous, associate dynamic contact behavior with interaction change trends, and obtain contact interaction trend characteristics.

[0020] The present invention is improved in that the step of obtaining the action mode conversion signal is specifically as follows:

[0021] S311: Based on the touch interaction trend characteristics, the identification state of the main control unit is judged, the current state is compared with the matching attributes of each control logic structure, and the correspondence between the identification state and the execution logic is judged in combination with the massage action type, control response rhythm and parameter settings to obtain the execution logic matching relationship.

[0022] S312: Based on the execution logic matching relationship, filter the control channel corresponding to the current massage action, analyze the parameter allocation mode in the control channel, calculate the correlation between the operation parameter configuration order and the logic path, adjust the allocation mode of each operation parameter, and obtain the parameter allocation configuration content.

[0023] S313: Based on the parameter allocation configuration content, determine the changes between the parameter allocation configuration content and the existing parameter content, analyze the conditions under which the switching criterion is established, combine the execution logic matching relationship and the parameter allocation adjustment direction, construct the action mode switching criterion, and obtain the action mode conversion signal.

[0024] The present invention is improved in that the step of obtaining the dynamic control instruction set is specifically as follows:

[0025] S411: Based on the motion mode conversion signal, determine the current operating state of the driver, collect the force change, analyze the continuous motion process during the motion, determine the spatial position change of the end massage component, and obtain the basic parameter group for drive control.

[0026] S412: Based on the aforementioned basic parameter set for drive control, compare the changes in force with the duration of motion, calculate the relative difference between the drive effect and the component inertia, analyze the spatial movement state, determine the offset trend of the contact area, and obtain the rhythm adjustment response factor.

[0027] S413: Based on the rhythm adjustment response factor, analyze the component status after the servo system rhythm switching, compare the motion rate and output synchronization during the action process, filter the sensitivity changes in the servo response, and obtain the dynamic control instruction set.

[0028] The present invention is improved in that the step of obtaining the synchronization deviation calibration result is specifically as follows:

[0029] S511: Based on the dynamic control instruction set, analyze the contact feedback data, compare the correspondence between the pressure sensor output and the spatial coordinates of the displacement detection unit, calculate the timing consistency of each contact point within the same action cycle, identify the data group that completes the time matching and position mapping, and obtain the periodic contact feedback data.

[0030] S512: Based on the periodic contact feedback data, analyze the corresponding items with the composite sensing state factor of the previous period, compare the pressure change direction of the same contact point in adjacent periods, and obtain the pressure gradient offset.

[0031] S513: Based on the pressure gradient offset, analyze the distribution of changes within a continuous cycle, compare the concentration of offset directions at different contact points, determine the deviation relationship between the current parameter state and the main control unit's discrimination standard, adjust the control parameters and discrimination standard, and obtain the synchronization deviation calibration result.

[0032] The present invention is improved in that the contact data refers to the pressure value and distribution information generated on the human body surface by each contact point in real time by the pressure sensor of the massage head or end massage component, and the pressure gradient distribution refers to the spatial variation trend and difference of the pressure magnitude at each point in the contact area.

[0033] A massage robot intelligent control system, the system comprising:

[0034] The sensing and acquisition module, based on the end-effector massage component, analyzes the contact data collected by the pressure sensor, combines the spatial coordinates output by the displacement detection unit, compares the movement direction with the pressure change trend, collects the dynamic response data of the acceleration detection unit, judges the pressure gradient and depth change at the contact point, and obtains the composite sensing state factor.

[0035] Based on the composite sensing state factor, the state discrimination module determines the pressure gradient distribution of each contact area, analyzes the change characteristics of the continuous spatial area, compares the relationship between the end movement direction and the contact depth change direction, identifies the dynamic contact state, and obtains the contact interaction trend characteristics.

[0036] The mode switching module adjusts the controller's response strategy to the interaction trend characteristics of the touch points, analyzes and identifies the state, changes the execution logic of the main control unit, filters the control channels that are compatible with the current massage action, adjusts the operation parameter allocation mode, uses the configuration parameter content as the switching basis, and obtains the action mode conversion signal.

[0037] Based on the motion mode conversion signal, the instruction output module adjusts the controller proportional adjustment configuration, synchronously refreshes the motion execution rhythm of the end massage component, switches to the new motion state, and obtains a dynamic control instruction set.

[0038] Based on the dynamic control instruction set, the calibration and adjustment module compares the contact feedback within the cycle, analyzes the gradient change of the composite sensing state factor in the previous cycle, determines the current deviation trend, adjusts the main control unit parameter settings and discrimination criteria, and obtains the synchronous deviation calibration result.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] In this invention, by constructing state criteria that are updated synchronously with the action cycle, pressure changes, spatial displacement, and dynamic response are incorporated into a unified analysis link, forming a continuous identification basis for the contact evolution process. By leveraging the correspondence between motion trends and force evolution, interactive trend identification is completed, enabling action switching to move beyond static triggering and instead be adjusted in an orderly manner around the current contact situation. Simultaneously, by combining rhythm reorganization and deviation review mechanisms, the control basis maintains convergence and correction capabilities during continuous operation, thereby enhancing the coordination of force distribution, the smoothness of action transition, and the stability of contact fit, and alleviating problems such as imbalance of light and heavy forces, abrupt switching, and slow compensation. Attached Figure Description

[0041] Figure 1 This is a flowchart of the main steps of the present invention;

[0042] Figure 2 This is a flowchart of the process for obtaining the composite sensing state factor in this invention;

[0043] Figure 3 This is a flowchart illustrating the process of obtaining touch interaction trend features in this invention.

[0044] Figure 4 This is a flowchart illustrating the acquisition of the action mode conversion signal in this invention.

[0045] Figure 5 This is a flowchart illustrating the acquisition of the dynamic control instruction set in this invention.

[0046] Figure 6 This is a flowchart illustrating the process of obtaining synchronization deviation calibration results in this invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0048] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0049] All user-related information involved in this invention (including but not limited to biometric information, identity verification information, behavioral data, device information, and other data that can be used for identity verification and personalized services) is collected and processed with the user's full knowledge and voluntary consent. The use of data is limited to purposes necessary for providing the technical services of this invention, and reasonable technical and management measures will be taken to ensure the security and confidentiality of users' personal information in terms of information protection and privacy.

[0050] Example

[0051] Please see Figure 1 This invention provides a technical solution, an intelligent control method for a massage robot, comprising the following steps:

[0052] S1: Based on the end massage component, analyze the contact data obtained by the pressure sensor, combine the spatial coordinates output by the displacement detection unit, compare the movement direction and the pressure change trend at each point, collect the dynamic response data of the acceleration detection unit, and determine the gradient distribution and depth change of the contact point pressure with the movement by performing periodic differential analysis on the contact data, and obtain the composite sensing state factor.

[0053] S2: Based on the composite sensing state factor, the pressure gradient distribution of each contact area is determined. By analyzing the change characteristics of the continuous spatial area, the relationship between the end motion direction and the contact depth change direction is compared. The consistency of direction is used as the criterion to identify the current dynamic contact state and obtain the contact interaction trend characteristics.

[0054] S3: Adjust the controller's response strategy to the trend characteristics of touch interaction, analyze and identify the state and change the execution logic of the main control unit. By filtering the control channel that is compatible with the current massage action, adjust the operation parameter allocation mode, and use the current configuration parameter content as the switching basis to obtain the action mode conversion signal.

[0055] S4: Based on the motion mode conversion signal, optimize the drive response mechanism, instruct the controller to adjust the proportional adjustment mode, synchronously refresh the motion execution rhythm of the end massage component, and switch the motion rhythm to the new state through the servo system to obtain a dynamic control instruction set;

[0056] S5: Based on the dynamic control instruction set, compare the contact feedback within the cycle, analyze the gradient change between the composite sensing state factor and the previous cycle, and adjust the main control unit parameter settings and discrimination criteria by judging the current deviation trend to obtain the synchronization deviation calibration result.

[0057] Composite sensing state factors include signal synchronization, sensing accuracy, and data fusion; contact interaction trend characteristics include action coordination, state recognition rate, and interaction continuity; action mode conversion signals include switching stability, drive adaptability, and parameter transmission consistency; dynamic control instruction set includes action response rate, output synchronization, and servo switching sensitivity; and synchronization deviation calibration results include deviation correction rate, calibration consistency, and adjustment accuracy.

[0058] In S1, contact data refers to the pressure values ​​and distribution information generated on the human body surface by each contact point, collected in real time by the pressure sensors of the massage head or end effector; the displacement detection unit refers to the sensor module installed at the end of the massage robot, used to measure the position (coordinates) and motion trajectory of the massage component in space, and can output three-dimensional coordinates; the motion direction refers to the instantaneous movement direction of the end effector during massage operation, usually obtained through the displacement detection unit, and is a vector reflecting the movement trend; the pressure change trend refers to the rise, fall, fluctuation, and other changes in pressure values ​​at one or more contact points over time and position, which can reveal the dynamics of force; the acceleration detection unit refers to the inertial sensor installed at the end effector. The system includes a motion measurement component that detects the acceleration of the massage component in real time, capturing the physical response of motion initiation, cessation, or speed changes; dynamic response data refers to the time-series data output by the acceleration detection unit that reflects changes in the motion of the end-effector (such as acceleration, deceleration, and turning); differential analysis, based on continuously acquired pressure data, analyzes the differences in pressure changes between adjacent moments within each cycle to identify the rate and direction of pressure changes; gradient distribution refers to the spatial distribution of pressure at each contact point, focusing on the increase or decrease of pressure from one point to another in space; and depth change refers to the change in the relative displacement between the massage component and the human body surface over time, reflecting the dynamic change process of massage depth.

[0059] In S2, pressure gradient distribution refers to the spatial variation trend and differences in pressure magnitude at various points within the contact area, focusing on the regional characteristics formed by local pressure increases or decreases; spatially continuous region refers to the area on the working surface of the massage robot where pressure or other sensor signals change continuously with spatial coordinates without abrupt changes, reflecting the overall trend of force; variation characteristics refer to the main laws or characteristic patterns exhibited when physical quantities such as pressure or displacement change with time and space within the above-mentioned areas; interrelationship refers to the directional relationship between the movement direction of the end massage component and the direction of contact depth change, used to determine whether the two are coordinated; dynamic contact state refers to the current state of dynamic interaction between the massage component and the skin surface, such as sliding and pushing, rather than static pressing.

[0060] In S3, the controller refers to the main control computing unit used to receive sensor data, perform logical judgments, and issue control signals, coordinating the work of various modules; the response strategy refers to the action execution rules selected and adjusted by the controller based on specific recognition results, including parameter settings and control mode switching; the main control unit refers to the core control module of the entire massage robot control system, which makes decisions and issues commands; the control channel refers to the logical path of different action modes or execution processes in the controller, such as dedicated logical channels for static pressing or dynamic sliding; the allocation mode refers to the allocation method of control parameters or execution tasks among various control channels, reflecting the scheduling of execution resources; the switching basis refers to the judgment criteria or signals by which the controller determines whether a mode switch is needed based on the currently recognized contact state and other data.

[0061] In S4, the drive response mechanism refers to the way the drive controller responds to commands issued by the main control unit, specifically manifested as the adjustment process of force, displacement, and other actions; the proportional adjustment method refers to the driver adjusting the driving force or speed proportionally when outputting control signals to the motor and other execution units to achieve a smooth transition of actions; the action execution rhythm refers to the time sequence characteristics of the speed, intensity, frequency, etc. of the massage component performing massage actions, which affects the massage effect and action synchronization.

[0062] In S5, contact feedback refers to the real-time feedback data generated by the interaction between the massage robot and the massaged area, which is detected by sensors as pressure, position, etc., during the action execution cycle; gradient change refers to the spatial distribution difference of physical quantities such as pressure between the current cycle and the previous cycle, used to detect changes in operation and force state; deviation trend refers to the analysis of the direction and magnitude of the change in feedback data between two cycles, reflecting whether the adjustment of action or parameters has led to the system state becoming more stable or changing; judgment criteria refer to the criteria used by the main control unit to determine whether the system state has reached the adjustment target, such as the smoothness of action or error range.

[0063] Please see Figure 2 The specific steps for obtaining the composite sensing state factors are as follows:

[0064] S111: Based on the end-effector massage component, analyze the contact data collected by the pressure sensor, monitor the spatial coordinate information output by the displacement detection unit, synchronously associate the pressure data with the spatial coordinates according to the time identifier, and construct the spatial position and pressure correspondence relationship by combining the actual distribution of the contact points to obtain the spatiotemporal pressure sequence of the contact points;

[0065] Based on the end-effector massage component, the sampling period of the pressure sensor is first fixed at 10 milliseconds, and the coordinate refresh period of the displacement detection unit is synchronously set to 10 milliseconds. Within the same sampling moment, the pressure values ​​and corresponding position data of each contact point are read. For example, at a certain moment, the pressures of three contact points are read as 12, 18, and 15, with corresponding positions of 10, 5, and 2; 12, 5, and 2; and 11, 6, and 2. This set of data is then uniformly labeled with the same time identifier and subsequently bound according to the contact point number to form a single-point correspondence record. The system continues to read data at the next moment, for example, if the pressure changes to 14, 20, and 16, and the position changes to 10.5, 5, and 2.1; and 12.5, 5... 1, 2.1, 11.5, 6, 2.1, and so on, are arranged in the same order to obtain the pressure and position changes of each contact point at continuous time. Then, the spatial distance between each contact point is detected, and points with a distance of no more than 3 mm are grouped into the same region. For example, if the distance between the first point and the second point is 2 mm, they are grouped into the same region, and the pressure values ​​of the current time in that region are arranged sequentially as 12, 18, 15. Then, the pressure values ​​of the corresponding region at the next time are arranged as 14, 20, 16. After that, the sequence is continued to be spliced ​​in chronological order, while retaining the spatial position label corresponding to each pressure value. Finally, a spatiotemporal pressure sequence of contact points is formed, which is continuously unfolded in time and arranged according to the position of the contact points.

[0066] S112: Based on the spatiotemporal pressure sequence of the contact point, determine the motion direction vector of the end massage component, combine the pressure distribution change trend over a continuous period, compare the correspondence between the two types of direction vectors by using the angle change method, associate the dynamic response information of the acceleration detection unit, and obtain the pressure direction consistency feature quantity.

[0067] The system acquires the end-effector position changes at two consecutive moments to determine the direction of movement within that time period. For example, if the position of a contact point changes from 10, 5, 2 to 10.5, 5, 2.1, then displacement occurs simultaneously along both the forward and downward directions. The system then reads the pressure changes at each contact point within the same time period. For instance, if three points change from 12, 18, 15 to 14, 20, 16, the corresponding increments are 2, 2, 1. The direction of pressure change is then determined based on the distribution of pressure increments within the contact area. Finally, the system compares the degree of deflection between the direction of movement and the direction of pressure change, dividing the deflection angle into three intervals: 0 to 30 degrees is considered consistent, 31 to 60 degrees is considered... The values ​​are defined as transitional, and values ​​between 61 and 180 are considered inconsistent. For example, if the comparison result is 20, it is classified as consistent. Then, the dynamic response values ​​output by the acceleration detection unit within the same time period are correlated. For example, if 0.8 is detected, the acceleration stability range is set to 0.5 to 1.5. Values ​​below 0.5 are considered insufficient change, and values ​​above 1.5 are considered too rapid change. After 0.8 falls into the stability range, it is screened together with the aforementioned direction consistency judgment. Only when the direction result is in the consistency range and the acceleration is in the stability range is the current time period recorded as valid consistency. After processing multiple consecutive time periods in sequence, the pressure direction consistency feature quantity arranged by time is obtained.

[0068] S113: Based on the consistent pressure direction characteristic, the continuous periodic pressure distribution is processed by periodic difference, and the pressure gradient change between spatially adjacent contact points and the contact depth change range are combined to organize and aggregate multiple types of sensing information to obtain the composite sensing state factor.

[0069] Pressure distribution data for the same area within two consecutive cycles are subtracted point by point. For example, if the previous cycle's data is 12, 18, 15, and the current cycle's is 14, 20, 16, the differences are 2, 2, and 1 respectively. Then, the data is grouped by proximity based on the contact point spacing, with points no more than 3 mm apart grouped together. The average difference within each group is calculated to obtain the pressure change level for that area during the current cycle. For example, the average result might be 1.67. Next, the pressure increase between adjacent points is examined. For instance, if the pressure difference between two adjacent points is 6 and the spacing is 2 mm, the change per millimeter is 3. This same process is repeated for all adjacent points within the area to obtain the overall gradient level for that cycle. Finally, the contact depth change is read. For example, if the end pressure position changes from 2 to 2.1, the depth change is 0.1. The average pressure difference, average gradient, and depth change are converted to a uniform scale according to preset ranges. The pressure difference range is set to 0 to 5, the gradient range to 0 to 4, and the depth change range to 0 to 1. After conversion, the values ​​are combined and calculated at 0.4, 0.4, and 0.2 to obtain a comprehensive result of 0.35. This result is then compared with the grading intervals: 0 to 0.3 is judged as low, 0.3 to 0.6 as medium, and above 0.6 as high. 0.35 falls into the medium level. Combined with the consistency mark of the corresponding cycle and the synchronization of the previous and subsequent cycles, a composite sensing state factor for subsequent control judgment is compiled.

[0070] Please see Figure 3 The specific steps for obtaining touch interaction trend characteristics are as follows:

[0071] S211: Based on the composite sensing state factor, analyze the pressure distribution of each contact point as a function of spatial coordinates, divide the region according to the spatial adjacency of the contact points, compare the direction of pressure change and the degree of continuity of distribution of adjacent contact points, and obtain the pressure gradient distribution characteristics.

[0072] The pressure values ​​and spatial coordinates of each contact point within the same sampling period are obtained. These contact points are then projected onto the working surface currently covered by the end-effector massage component in numerical order. For example, at a certain moment, the pressure values ​​of the five contact points are 11, 14, 18, 17, and 13, corresponding to lateral positions of 2, 4, 6, 8, and 10, while maintaining consistent longitudinal positions. First, the lateral adjacent distance is checked, and points with a spacing of no more than 3 mm are grouped into the same area. Then, the pressure difference between every two adjacent points is calculated: the difference between the first two points is 3, between the middle two points is 4, and between the last two points is 1. The positive and negative directions of the difference are recorded: if the latter point is greater than the former, it is recorded as a positive increase; if the latter point is less than the former, it is recorded as a negative decrease. This allows for the calculation of the continuous increase in the front section and the decrease in the rear section. The process involves observing the decreasing changes in pressure within each segment, and then checking the continuity between three adjacent points. A continuous segment is defined as two consecutive differences in the same direction with a fluctuation range not exceeding 5. A discontinuous segment is defined as one where the direction changes and the difference jump exceeds 5. For example, if the difference increases continuously from 11 to 14 and then to 18 with jumps of 3 and 4 respectively, it is classified as a continuous segment. If the difference changes in the opposite direction from 18 to 17, a boundary is formed at that point. The pressure arrangement within each region is then organized, and increasing, decreasing, and peak segments are labeled. Finally, the contact depth records for that period (0.6, 0.7, 0.9, 0.8, 0.6) are used to confirm the location of the pressure center corresponding to the peak point, thus forming the pressure gradient distribution characteristics within the current contact region.

[0073] S212: Based on the pressure gradient distribution characteristics, track the changing trend of each continuous spatial region on the time axis, calculate the pressure distribution change direction of each continuous region, compare the vector correspondence between the end motion direction and the contact depth change direction, and obtain the directional coordination relationship characteristics.

[0074] The pressure distribution of the same region over three consecutive sampling periods is sequentially expanded. For example, the first region shows pressure values ​​of 11, 14, and 18 in the first period, 12, 16, and 19 in the second period, and 13, 17, and 20 in the third period. The position of the highest pressure point is checked periodically to ensure it remains in the center of the end region. The difference between the beginning and end of each period is calculated: 7 for the first period, 7 for the second, and 7 for the third. Regions where the difference between the two pressure values ​​does not exceed 2 and the highest point moves no more than 2 millimeters within consecutive periods are designated as stable continuous regions. The end displacement records for the corresponding time period within this region are then read. For example, if the end moves forward laterally by 0.5, 0.6, and 0.5 mm in the three periods, while simultaneously downward by 0.1 and 0 mm... 0.1 and 0.2 are organized as forward movement accompanied by continuous downward pressure. The direction of contact depth change is then read. If the depth value increases from 0.6 to 0.8 and then to 0.9, it is recorded as continuous deepening. The forward movement direction and the depth deepening direction are compared side by side. The direction coordination judgment is divided into three levels: deviation less than 30 is recorded as high coordination, 30 to 60 is recorded as medium coordination, and greater than 60 is recorded as low coordination. In this example, the movement direction and the depth change direction are both advancing in the same direction in the continuous cycle and the deviation records are 18, 22 and 25 respectively, all falling into the high coordination range. Combined with the record that the pressure peak position in the region did not jump across regions, the continuous spatial region is marked as the continuous segment of direction coordination, and finally the direction coordination relationship characteristics are obtained.

[0075] S213: Based on the directional coordination relationship characteristics, compare the synchronization between the pressure gradient distribution change and the contact depth change in the contact area, identify the contact segments where the directional relationship remains continuous, associate dynamic contact behavior with the interaction change trend, and obtain the contact interaction trend characteristics.

[0076] Within the same contact area, retrieve the pressure gradient change records and contact depth change records for two consecutive cycles. For example, if the adjacent pressure differences in a certain area in the previous cycle were 3, 4, and 1, and in the current cycle they become 4, 4, and 2, while the contact depth changes from 0.7 to 0.9, first compare the synchronous direction of the pressure gradient change and the depth change. If the overall gradient increases and the depth increases synchronously, it is recorded as a synchronous increase; if the gradient decreases and the depth increases, it is recorded as a reverse change. In this example, the total gradient increases from 8 to 10, and the depth increases from 0.7 to 0.9, so it is first recorded as a synchronous increase. Then, perform the same check on three consecutive cycles. If the synchronous markers reach more than two times within a consecutive cycle, the area is listed as a continuous contact segment. Subsequently, check whether the directional relationship remains continuous, and then... The motion direction record, pressure peak migration direction record, and depth change direction record are aligned item by item. For example, if all three directions point forward and downward in a continuous cycle and the deviation record is controlled within 25, then the direction is considered to be continuous. If the deviation of any direction exceeds 60, then the cycle is removed from the continuous segment. Then, the dynamic contact behavior record is associated, and the three types of behavior, sliding, pushing, and pausing, are expressed in the order of the cycle. For example, the first two cycles are sliding and pressing down, and the third cycle is a short pause. The interaction change trend is then classified. Continuous sliding accompanied by increasing depth is recorded as the enhancement segment, sliding weakens and the depth remains flat as the stable segment, and pausing and gradient decline is recorded as the retreat segment. In this example, the first two cycles are classified as the enhancement segment, and the third cycle is classified as the stable segment. Finally, the contact interaction trend characteristics are obtained.

[0077] Please see Figure 4 The specific steps for obtaining the action mode switching signal are as follows:

[0078] S311: Based on the trend characteristics of touch interaction, the identification status of the main control unit is judged, the current status is compared with the matching attributes of each control logic structure, and the correspondence between the identification status and the execution logic is determined by combining the massage action type, control response rhythm and parameter settings to obtain the execution logic matching relationship;

[0079] The main control unit retrieves the interaction classification results, directional continuity markers, pressure gradient levels, and contact depth levels for the current cycle, and then writes the state values ​​into the current state cache. For example, the current record is the enhanced segment, the directional continuity marker is 1, the pressure level is medium-high, and the depth level is medium. At this time, it continues to read the preset control logic table in the main control unit, unfolding the three types of logic—static pressing, vertical pushing, and rhythm switching—in sequence, and retrieving the corresponding action tags, rhythm settings, pressure ranges, and depth ranges for each item. Then, it compares the current state with each logic item by item. 1 point is awarded for a matching action tag, 1 point for a rhythm difference of no more than 0.5, 1 point for a pressure value falling into the corresponding range, and 1 point for a depth value falling into the corresponding range. Each corresponding interval is scored as 1 point. For example, if the current action is vertical movement, the rhythm is 1.8 times per second, the average pressure is 16, and the contact depth is 0.8, then when the action is consistent with the vertical movement logic, it scores 1 point; the difference between the rhythm and the preset 2 times per second is 0.2, it scores 1 point; the pressure of 16 falls into the 12 to 20 interval, it scores 1 point; and the depth of 0.8 falls into the 0.6 to 1.0 interval, it scores 1 point. The total score is 4. After comparing with the other logics in the same order, 0 to 1 is recorded as low match, 2 to 3 as medium match, and 4 as high match. The logic number with the highest score and high match, the corresponding parameter interval, and the rhythm setting are registered as the corresponding item of the current recognition state, thus obtaining the execution logic matching relationship.

[0080] S312: Based on the execution logic matching relationship, filter the control channels corresponding to the current massage action, analyze the parameter allocation mode within the control channel, calculate the correlation between the operation parameter configuration order and the logic path, adjust the allocation mode of each operation parameter, and obtain the parameter allocation configuration content.

[0081] The system reads the candidate channels corresponding to the high-match logic from the channel table within the controller. If the current high-match result corresponds to vertical shifting, it sequentially retrieves the action labels, rhythm upper limits, parameter order, and logic node tables of shifting channel A, shifting channel B, and composite channel C. Then, it checks the current massage action label against the usage labels of each channel item by item. Channels with matching action labels are retained, while those with inconsistent labels are discarded. For example, if the current action label is vertical shifting, A and C are retained first. Then, the current rhythm of 1.8 times per second is compared with the channel's allowed range. A corresponds to 1.2 to 2.2, and C corresponds to 0.8 to 1.6. Therefore, A is retained, and C is discarded. Next, the current parameter order within channel A is read. Let's assume the original order is strength 16, speed 2... 4. Displacement 6, pause 0.2. Then retrieve the logic node requirements for this channel. If the node order is displacement, speed, force, pause, check whether the current position corresponds to the correct node. If the displacement is currently in the third position, record 1 misplacement. If the speed is currently in the second position, record 1 correct position. If the force is currently in the first position, record 1 misplacement. If the pause is currently in the fourth position, record 2 correct positions. Accumulate 2 misplacements. Then record 0 as high association, 1 to 2 as medium association, and 3 and above as low association. After determining that the current order is medium association, perform rearrangement, placing displacement 6 in the first position, speed 24 in the second position, force 16 in the third position, and pause 0.2 in the last position. Write the rearranged parameter queue into the execution cache to obtain the parameter allocation configuration content.

[0082] S313: Based on the parameter allocation configuration content, determine the changes between the parameter allocation configuration content and the existing parameter content, analyze the conditions under which the switching criterion is established, combine the execution logic matching relationship and the parameter allocation adjustment direction, construct the action mode switching criterion, and obtain the action mode conversion signal;

[0083] Retrieve the new parameter queue for the current cycle, then retrieve the existing parameter queue from the previous cycle. For example, if the current queue contains displacement 6, velocity 24, force 16, and pause 0.2, and the previous cycle contained displacement 4, velocity 20, force 18, and pause 0.3, calculate the change for each item: displacement changes by 2, velocity by 4, force by 2, and pause by 0.1. Compare each change with a preset range: displacement change less than 1 is recorded as a fine adjustment, 1 to 3 as a medium adjustment, and greater than 3 as an emphasis; velocity change less than 2 is recorded as a fine adjustment, 2 to 5 as a medium adjustment, and greater than 5 as an emphasis; force change less than 2 is recorded as a fine adjustment, 2 to 4 as a medium adjustment, and greater than 4 as an emphasis; pause change less than 0.05 is recorded as a minor adjustment. The adjustment range is 0.05 to 0.15, which is considered medium adjustment, and greater than 0.15 is considered strong adjustment. In this example, all four items fall within the medium adjustment range. Then, the two conditions of the switching criterion are checked: first, whether the execution logic matching level reaches high matching; and second, whether the parameter adjustment direction is consistent with the current interaction trend. If the current trend is an enhancement segment, it is required that the displacement and speed do not decrease and the pause does not increase. In this example, the displacement increases from 4 to 6, the speed increases from 20 to 24, and the pause decreases from 0.3 to 0.2, which meets the direction consistency condition. Then, the high matching flag and the direction consistency flag are set to true at the same time. When both conditions are true at the same time, the target mode number, target channel number, and parameter adjustment level sequence are written to obtain the action mode conversion signal.

[0084] Please see Figure 5 The specific steps for obtaining the dynamic control instruction set are as follows:

[0085] S411: Based on the motion mode conversion signal, determine the current operating status of the driver, collect the force change, analyze the continuous motion process during the motion, determine the spatial position change of the end massage component, and obtain the basic parameter group for drive control.

[0086] The driver reads the current target mode number, target channel number, and parameter adjustment level sequence. Simultaneously, it retrieves the operating gear, motor output current, actuator displacement record, and end-effector pressure feedback record from the driver's real-time status register. The current operating state is divided into three intervals: pending switching, switching in progress, and switched. A driver output change of less than 5 and a displacement change of less than 0.5 is considered pending switching; a driver output change between 5 and 15 and a displacement change between 0.5 and 2 is considered switching in progress; and a driver output change greater than 15 with a stable displacement direction for two consecutive sampling periods is considered switched. For example, if the current current increases from 1.2 to 1.8 (a change of 0.6), after conversion to the preset drive response level, it falls into the switching in progress interval. The driver then continues to collect force changes, reading the force values ​​at three consecutive sampling points of the end-effector massage component. If these values ​​are 14, 16, and 17 respectively, the difference between adjacent points is calculated to obtain the desired result. 2 and 1, then the difference change is recorded as stable if less than 1, gradual if 1 to 3, and abrupt if greater than 3. In this example, the change in force is recorded as gradual. Then, the position records of the end at the three sampling times are read along the direction of the action. For example, the coordinates are 10, 5, 2, 10.8, 5, 2.2, 11.5, 5, 2.4 respectively. Then the cumulative lateral displacement is 1.5 and the cumulative longitudinal downward pressure is 0.4. Then, check whether there is a direction reversal during the continuous movement. If the lateral displacement of two adjacent sampling segments is positive and the longitudinal displacement is positive, it is recorded as continuous advancement. If there is a positive-negative switch in any direction, it is recorded as interrupted advancement. In this example, the three consecutive sampling points are moving forward and downward, so it is recorded as continuous advancement. Then, the current driving state, the level of force change, the cumulative value of lateral displacement, the cumulative value of longitudinal displacement, and the continuous movement mark are written into the control cache of this cycle. Finally, the basic parameter group of driving control is obtained.

[0087] S412: Based on the fundamental parameter set of drive control, compare the changes in force with the duration of motion, calculate the relative difference between the drive effect and the component inertia, analyze the spatial movement state, determine the offset trend of the contact area, and use the following formula:

[0088] ;

[0089] The rhythm regulation response factor is obtained, where, The rhythm regulation response factor is used to characterize the overall response level during the control rhythm adjustment process. The driving force ratio refers to the ratio of the current driving force to the preset reference force. The runtime ratio represents the ratio of the current duration of an action to a preset reference duration. The inertia-impulse ratio represents the ratio of the impulse term related to the inertia of the end-effector to the preset reference impulse term. The spatial displacement ratio refers to the ratio of the spatial movement distance to the preset reference movement distance. The contact offset ratio is the ratio of the spatial contact distribution offset amplitude to the preset reference offset amplitude.

[0090] The rhythm adjustment response factor is a numerical parameter obtained by comprehensively normalizing key parameters such as drive, inertia, spatial movement and contact offset. It is used to quantify and characterize the response characteristics of the controller to the motion state of the end massage component during rhythm adjustment. It reflects the comprehensive response intensity and coordination of the control system when adjusting the action rhythm under the combined action of the current driving force, motion duration and component inertia, combined with changes in spatial displacement and contact offset.

[0091] Comparing the changes in force with the duration of motion, continuous periodic samples of the massage robot's end effector were selected. The force in period one was collected from the end effector drive motor tension sensor, with a recorded value of 38N and a reference value of 40N. The extreme value normalization method was used to derive the corresponding driving force ratio. Meanwhile, the running duration within this cycle is recorded as 2.2s by the main control unit's time counting module, and the corresponding reference running time is set to 2s. After normalization in the same way, the running time ratio is obtained. The component's inertia impulse was obtained by integrating the angular acceleration data recorded by the end effector module. The measured impulse value was 0.084 kg·m² / s, and the reference impulse was set to 0.1 kg·m² / s. The normalized inertia impulse ratio was... The spatial movement distance, measured by the three-dimensional displacement detection module installed at the end, was 0.24m, with a reference spatial distance of 0.2m. The normalized distance was... The contact offset amplitude is recorded by an array of multi-point pressure sensors, showing the range of their offset vector distribution. The maximum offset spacing in the current cycle is 0.004m, and the reference offset amplitude is 0.01m. After normalization, the contact offset ratio is obtained. Substitute the normalized parameters into the rhythm regulation response factor formula: The response factor interval is divided into three segments: when When this occurs, it indicates that the current cycle system is in a low-response rhythm phase, and the rhythm switching tends to be stable; when When this indicates the entry into the medium response rhythm phase, the control system may make moderate rhythm adjustments; when When this occurs, it is determined to be a high-response rhythm segment, requiring significant rhythm refresh and controller adjustment actions. The current value is... Since the system is in a low-response rhythm range, it can be inferred that the dynamic balance formed by the drive input and structural inertia is relatively stable within this cycle, the impact of external disturbances is small, and the system rhythm does not need to be reconfigured. Based on this judgment, the main control unit will register it as a stable signal and pass the status label to the next control flow as the input basis for screening servo sensitivity change strategies.

[0092] S413: Based on the rhythm adjustment response factor, analyze the component status after the rhythm switching of the servo system, compare the motion rate and output synchronization during the action process, filter the sensitivity changes in the servo response, and obtain the dynamic control instruction set.

[0093] The servo system reads the current motion frequency, target motion frequency, output pulse interval, and end-effector displacement record after the rhythm switch. Then, it categorizes the component states: a difference of less than 0.2 between the target frequency and the current frequency is recorded as a stable range; 0.2 to 0.5 is recorded as a transition range; and greater than 0.5 is recorded as a deviation range. For example, if the target frequency is 2 times per second and the current frequency is 1.7 times per second, the difference is 0.3, which is initially recorded as a transition range. Next, the motion rate during the motion process is continuously read. Assuming the displacement increments in three consecutive sampling periods are 0.9, 1.0, and 1.1, with corresponding time intervals of 0.05, the converted motion rates are 18, 20, and 22 respectively. Then, the drive output record for the same time period is retrieved. If the theoretical rates corresponding to the output beats are 19, 20, and 21 respectively, the actual rates are compared with the output beats item by item. A difference less than 1 is recorded as high synchronization; a difference between 1 and 21 is recorded as high synchronization. A value of 3 is considered medium synchronization, and values ​​greater than 3 are considered low synchronization. In this example, the differences between the three sampling segments are 1, 0, and 1, respectively, so the overall result is considered high synchronization. Subsequently, the sensitivity changes in the servo response are screened, and the response delay before and after the rhythm switch is read. If the response delay before the switch is 0.12 and after the switch it is 0.08, then the change is 0.04. A delay reduction of less than 0.02 is recorded as low change, 0.02 to 0.05 is recorded as medium change, and greater than 0.05 is recorded as high change. In this example, it is recorded as medium change. Then, the pressure feedback fluctuation value within two consecutive cycles after the switch is read. If the fluctuation value is 1.5 and 1.8 respectively, and neither exceeds the preset threshold of 3, then the current sensitivity level is retained. If it exceeds 3, it is downgraded by one level. Finally, the frequency difference range, motion rate synchronization level, sensitivity change level, and component status flag after the switch are written into the control instruction cache in sequence to form a dynamic control instruction set.

[0094] Please see Figure 6 The specific steps for obtaining the synchronization deviation calibration results are as follows:

[0095] S511: Based on the dynamic control instruction set, analyze the contact feedback data, compare the correspondence between the pressure sensor output and the spatial coordinates of the displacement detection unit, calculate the timing consistency of each contact point within the same action cycle, identify the data group that completes the time matching and position mapping, and obtain the periodic contact feedback data.

[0096] The pressure feedback and spatial position records of the end-effector massage component within the current action cycle are retrieved and sorted according to a uniform sampling interval. Assuming five consecutive samples are taken within the same cycle, the pressure sensor outputs are 13, 15, 16, 15, and 14, respectively. The corresponding lateral positions of the displacement detection unit are 10, 11, 12, 13, and 14, respectively, and the vertical downward pressure positions are 2.0, 2.2, 2.4, 2.3, and 2.1, respectively. First, the time markers of each sampling point are checked for one-to-one correspondence. If the time deviation between two adjacent data sets is less than 5 milliseconds, it is considered a time match; 5 to 10 milliseconds is considered a weak match; and greater than 10 milliseconds is directly discarded. In this example, the time deviations of the five data sets are 2, 3, 2, 4, and 3, all of which are considered time matches. Then, the pressure value and spatial position are checked point by point to ensure a stable mapping. First, the increasing lateral position and synchronously rising pressure are used as the initial effective mapping condition, followed by the downward pressure position... The condition for effective mapping in the subsequent stage is that the pressure decreases synchronously during the retraction. In this example, the lateral position of the first three sampling points increases from 10 to 12, the pressure increases from 13 to 16, and the downward pressure position increases from 2.0 to 2.4, which is considered as valid mapping in the first stage. The lateral position of the last two sampling points continues to increase, but the downward pressure position decreases from 2.4 to 2.1, and the pressure decreases from 16 to 14, which is considered as valid mapping in the retraction stage. Then, a timing consistency check is performed on each contact point within the same action cycle. If the time difference between the pressure peaks of adjacent contact points does not exceed one sampling interval, it is recorded as high consistency. If the difference is two sampling intervals, it is recorded as medium consistency. If the difference is three or more sampling intervals, it is recorded as low consistency. For example, if the peaks of the three contact points appear in the 3rd, 3rd, and 4th samplings, it is judged as high consistency. Finally, only the sampling groups that simultaneously satisfy time matching and position mapping are retained, and the data retained in this cycle are reorganized according to the contact point number, time order, and spatial coordinate order to obtain the cycle contact feedback data.

[0097] S512: Based on periodic contact feedback data, analyze the corresponding terms of the composite sensing state factor of the previous period, compare the pressure change direction of the same contact point in adjacent periods, and use the formula:

[0098] ;

[0099] The pressure gradient offset is obtained, where, This represents the pressure gradient offset at the nth contact point. This represents the pressure measurement result at the nth contact point within the current cycle. This represents the pressure measurement result at the nth contact point within the previous cycle. The parameters representing the consistency quantization of displacement direction at the current position. Indicates the length of the contact feature within the current period. This represents the normalized parameter of the dynamic amplitude of the acceleration at the nth contact point within the current cycle. This means adding 1 Find the square root, used for dynamic normalization.

[0100] Pressure gradient offset ( Pressure gradient offset (PRA) refers to a parameter calculated after normalization for each contact point during the operation of a massage robot. It compares the pressure measurement results of the current cycle with the previous cycle, combines the displacement direction changes and dynamic amplitude of the end effector components, and then normalizes the results. PRA reflects the extent to which the pressure distribution in a local area of ​​each contact point has changed compared to the previous cycle during dynamic movement and contact adjustment of the robot's end effector. It incorporates spatial displacement direction and dynamic motion factors as correction terms, ensuring that this indicator reflects both the temporal change in pressure and the correlation between spatial and dynamic states.

[0101] Extract the raw pressure sensor readings for the same contact point within two consecutive operating cycles. Taking contact points P1, P2, and P3 as examples, the pressure measurement value for the current cycle is... The pressure values ​​are 18.2 kPa, 21.5 kPa, and 19.6 kPa, respectively, corresponding to the pressure values ​​of the previous cycle. The pressures were 16.4 kPa, 20.2 kPa, and 19.0 kPa, respectively. The calculated pressure changes were 1.8 kPa, 1.3 kPa, and 0.6 kPa. The consistency parameters of the displacement direction at the current location were analyzed. At that time, the cosine method of the angle between the displacement vectors within the period was used for quantization, and the original consistency coefficients of the three points were obtained as 0.94, 0.89, and 0.91, respectively. The dynamic amplitude of acceleration was then analyzed. At that time, the peak acceleration values ​​of the end effector during the current cycle were collected as 1.5 m / s², 2 m / s², and 1.8 m / s², respectively. These values ​​were normalized using the maximum acceleration value of 3 m / s² to obtain the corresponding values. The normalized results were 0.5, 0.67, and 0.6, respectively, and the contact characteristic length was analyzed. At that time, the actual contact surface diameter was recorded as 24mm, which was uniformly converted to 0.024m, and the contact surface diameter was used as the characteristic length; all parameters were substituted into the formula, where, This indicates the change in pressure between the current period and the previous period. As a directional consistency factor, For contact feature length, For the dynamic normalization coefficients, perform independent calculations for each term and obtain the results: For P1, substitute: , , , ,get: For P2, substitute: , , , ,get: For P3, substitute: , , , ,get: ;Preset pressure gradient offset determination range, when When this is classified as a severe offset interval, it indicates that the contact point experiences significant pressure changes between adjacent cycles, accompanied by strong inconsistencies in displacement direction; when When the value is classified as being in the moderate offset range, it indicates that there is a relatively obvious trend of mechanical change at the contact point, but it has not yet broken through the critical stability boundary; when When this is the case, it is classified as a slight deviation range, indicating that the cyclic pressure and displacement coordination at the contact point remains within an acceptable range, with only minor disturbances; when When the condition is classified as a stable interval, it indicates that the contact point is in a highly stable state with no significant fluctuations in mechanical parameters during the cycle. Substituting the calculated results into the above interval judgment rules for corresponding analysis: the pressure gradient offset at contact point P1 is... kPa, satisfying It is in the severely offset range; the pressure gradient offset at contact point P2 is kPa, satisfying It is in the medium offset range; the pressure gradient offset at contact point P3 is kPa, satisfying It is in the range of slight deviation.

[0102] The result shows that the calculated It reflects the degree of coupling between the force state change and displacement direction of the contact point during continuous operation cycle. The specific range of its value determines the current dynamic stability level of the contact point. Different ranges represent different degrees of operational instability. This quantity is used as an output result to drive the synchronization deviation calibration logic in the next step.

[0103] S513: Based on the pressure gradient offset, analyze the distribution of changes in a continuous cycle, compare the concentration of offset directions at different contact points, determine the deviation relationship between the current parameter state and the main control unit's discrimination standard, adjust the control parameters and discrimination standard, and obtain the synchronization deviation calibration result.

[0104] Acquire the gradient offset records of each contact point within three consecutive action cycles, and expand them sequentially according to the contact point number. Let the offset values ​​of the first cycle be 1.2, 1.5, and 1.4; the second cycle be 1.8, 2.0, and 1.9; and the third cycle be 2.1, 2.3, and 2.2. First, compare the offset change direction of the same contact point within consecutive cycles. If it continuously increases, it is recorded as positive concentration; if it continuously decreases, it is recorded as negative concentration; if it increases first and then decreases, or decreases first and then increases, it is recorded as dispersed change. In this example, all three contact points are continuously increasing, so they are first recorded as positive concentration. Then, check the offset difference between different contact points within the same cycle. If the maximum difference is less than 0.3, it is recorded as high concentration; 0.3 to 0.6, it is recorded as medium concentration; and greater than 0.6, it is recorded as low concentration. For example, the difference between the maximum value of 2.3 and the minimum value of 2.1 in the third cycle is 0.2, so it is recorded as high concentration. Then, compare the current parameter state with the discrimination criteria in the main control unit item by item. Let the current parameter state be... The forward pressure control value is 17, the speed control value is 23, and the downward pressure control value is 0.9. The standard ranges corresponding to the main control unit are 14 to 18, 20 to 24, and 0.7 to 1.0, respectively. All three parameters fall within the standard range. Further checks are made to see if the offset concentration direction is consistent with the current discrimination direction of the main control unit. If the main control unit requires maintaining a positive increase within the enhanced section and the concentration level is not lower than the medium concentration, then the current state meets the discrimination conditions. If the offset concentration direction is consistent but the average offset exceeds 2.5, then the pressure control value is reduced by 1 level, the speed control value is reduced by 1 level, and the downward pressure control value is reduced by 0.1. In this example, the average offset in the third cycle is 2.2, which does not exceed 2.5. Therefore, the pressure is kept unchanged at 17, the speed at 23, and the downward pressure at 0.9. Only the concentration threshold in the main control unit is tightened from 0.5 to 0.4, and the number of consecutive increasing judgment cycles is increased from 2 to 3. Finally, the synchronization deviation calibration result is obtained.

[0105] A massage robot intelligent control system, the system includes:

[0106] The sensing and acquisition module, based on the end-effector massage component, analyzes the contact data collected by the pressure sensor, combines the spatial coordinates output by the displacement detection unit, compares the movement direction with the pressure change trend, collects the dynamic response data of the acceleration detection unit, judges the pressure gradient and depth change at the contact point, and obtains the composite sensing state factor.

[0107] The state discrimination module is based on composite sensing state factors to determine the pressure gradient distribution of each contact area, analyze the change characteristics of the spatial continuous area, compare the relationship between the end motion direction and the contact depth change direction, identify the dynamic contact state, and obtain the contact interaction trend characteristics.

[0108] The mode switching module adjusts the controller's response strategy to the trend characteristics of touch interaction, analyzes and identifies the state, changes the execution logic of the main control unit, filters the control channel that is compatible with the current massage action, adjusts the operation parameter allocation mode, uses the configuration parameter content as the basis for switching, and obtains the action mode conversion signal.

[0109] Based on the motion mode conversion signal, the instruction output module adjusts the controller's proportional adjustment configuration, synchronously refreshes the motion execution rhythm of the end massage component, switches to the new motion state, and obtains a dynamic control instruction set.

[0110] The calibration and adjustment module is based on a dynamic control instruction set. It compares the contact feedback within the cycle, analyzes the gradient change of the composite sensing state factor with that of the previous cycle, determines the current deviation trend, adjusts the main control unit parameter settings and discrimination criteria, and obtains the synchronous deviation calibration result.

[0111] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for intelligent control of a massage robot, characterized in that, Includes the following steps: S1: Based on the end-effector massage component, analyze the contact data collected by the pressure sensor, compare the movement direction and pressure change trend, perform periodic difference analysis on the contact data, extract pressure gradient and depth change features, and obtain composite sensing state factors. S2: Based on the composite sensing state factor, analyze the pressure gradient and continuous change characteristics of each contact area, compare the direction of motion with the direction of contact depth change, use the consistency of direction to determine the dynamic contact state, and obtain the contact interaction trend characteristics. S3: Adjust the controller's response strategy to the touch interaction trend characteristics, combine the recognition state change execution logic, filter the control path that matches the current massage action, adjust the operation parameter allocation mode, and obtain the action mode conversion signal; S4: Based on the action mode conversion signal, adjust the drive response relationship, update the action execution rhythm and ratio adjustment mode, switch to the new action state, and obtain a dynamic control instruction set; S5: Based on the dynamic control instruction set, compare the contact feedback within the cycle, analyze the gradient change of the composite sensing state factor in the previous cycle, determine the current deviation trend, adjust the main control unit parameter settings and discrimination criteria, and obtain the synchronization deviation calibration result.

2. The intelligent control method for a massage robot according to claim 1, characterized in that, The composite sensing state factor is a quantitative indicator calculated based on signal synchronization, sensing accuracy, and data fusion. The touch interaction trend characteristics include action coordination, state recognition rate, and interaction continuity. The action mode conversion signal includes switching stability, drive adaptability, and parameter transmission consistency. The dynamic control instruction set includes action response rate, output synchronization, and servo switching sensitivity. The synchronization deviation calibration result includes deviation correction rate, calibration consistency, and adjustment accuracy.

3. The intelligent control method for a massage robot according to claim 1, characterized in that, The specific steps for obtaining the composite sensing state factor are as follows: S111: Based on the end-effector massage component, analyze the contact data collected by the pressure sensor, monitor the spatial coordinate information output by the displacement detection unit, synchronously associate the pressure data with the spatial coordinates according to the time identifier, and construct the spatial position and pressure correspondence relationship by combining the actual distribution of the contact points to obtain the spatiotemporal pressure sequence of the contact points; S112: Based on the spatiotemporal pressure sequence of the contact point, determine the motion direction vector of the end massage component, combine the pressure distribution change trend over a continuous period, compare the correspondence between the two types of direction vectors using the angle change method, associate the dynamic response information of the acceleration detection unit, and obtain the pressure direction consistency feature quantity. S113: Based on the pressure direction consistency feature quantity, the continuous periodic pressure distribution is processed by periodic differential processing, and the pressure gradient change between spatially adjacent contact points and the contact depth change range are combined to organize and aggregate multiple types of sensing information to obtain a composite sensing state factor.

4. The intelligent control method for a massage robot according to claim 1, characterized in that, The specific steps for obtaining the touch interaction trend features are as follows: S211: Based on the composite sensing state factor, analyze the pressure distribution of each contact point as a function of spatial coordinates, divide the region according to the spatial adjacency of the contact points, compare the direction of pressure change and the degree of continuity of distribution of adjacent contact points, and obtain the pressure gradient distribution characteristics. S212: Based on the pressure gradient distribution characteristics, track the changing trend of each continuous spatial region on the time axis, calculate the changing direction of pressure distribution in each continuous region, compare the vector correspondence between the end motion direction and the contact depth change direction, and obtain the directional coordination relationship characteristics. S213: Based on the directional coordination relationship characteristics, compare the synchronization between the pressure gradient distribution change and the contact depth change within the contact area, identify contact segments where the directional relationship remains continuous, associate dynamic contact behavior with interaction change trends, and obtain contact interaction trend characteristics.

5. The intelligent control method for a massage robot according to claim 1, characterized in that, The specific steps for obtaining the action mode switching signal are as follows: S311: Based on the touch interaction trend characteristics, the identification state of the main control unit is judged, the current state is compared with the matching attributes of each control logic structure, and the correspondence between the identification state and the execution logic is judged in combination with the massage action type, control response rhythm and parameter settings to obtain the execution logic matching relationship. S312: Based on the execution logic matching relationship, filter the control channel corresponding to the current massage action, analyze the parameter allocation mode in the control channel, calculate the correlation between the operation parameter configuration order and the logic path, adjust the allocation mode of each operation parameter, and obtain the parameter allocation configuration content. S313: Based on the parameter allocation configuration content, determine the changes between the parameter allocation configuration content and the existing parameter content, analyze the conditions under which the switching criterion is established, combine the execution logic matching relationship and the parameter allocation adjustment direction, construct the action mode switching criterion, and obtain the action mode conversion signal.

6. The intelligent control method for a massage robot according to claim 1, characterized in that, The specific steps for obtaining the dynamic control instruction set are as follows: S411: Based on the motion mode conversion signal, determine the current operating state of the driver, collect the force change, analyze the continuous motion process during the motion, determine the spatial position change of the end massage component, and obtain the basic parameter group for drive control. S412: Based on the aforementioned basic parameter set for drive control, compare the changes in force with the duration of motion, calculate the relative difference between the drive effect and the component inertia, analyze the spatial movement state, determine the offset trend of the contact area, and obtain the rhythm adjustment response factor. S413: Based on the rhythm adjustment response factor, analyze the component status after the servo system rhythm switching, compare the motion rate and output synchronization during the action process, filter the sensitivity changes in the servo response, and obtain the dynamic control instruction set.

7. The intelligent control method for a massage robot according to claim 1, characterized in that, The specific steps for obtaining the synchronization deviation calibration results are as follows: S511: Based on the dynamic control instruction set, analyze the contact feedback data, compare the correspondence between the pressure sensor output and the spatial coordinates of the displacement detection unit, calculate the timing consistency of each contact point within the same action cycle, identify the data group that completes the time matching and position mapping, and obtain the periodic contact feedback data. S512: Based on the periodic contact feedback data, analyze the corresponding items with the composite sensing state factor of the previous period, compare the pressure change direction of the same contact point in adjacent periods, and obtain the pressure gradient offset. S513: Based on the pressure gradient offset, analyze the distribution of changes within a continuous period, compare the concentration of offset directions at different contact points, determine the deviation relationship between the current parameter state and the main control unit's discrimination standard, adjust the control parameters and discrimination standard, and obtain the synchronization deviation calibration result.

8. The intelligent control method for a massage robot according to claim 1, characterized in that, The contact data refers to the pressure values ​​and distribution information generated on the human body surface by each contact point in real time, which are collected by the pressure sensor of the massage head or end massage component. The pressure gradient distribution refers to the spatial trend and difference of the pressure magnitude at each point in the contact area.

9. A smart control system for a massage robot, characterized in that, The system is used to implement the intelligent control method for the massage robot according to any one of claims 1-8, and the system includes: The sensing and acquisition module, based on the end-effector massage component, analyzes the contact data collected by the pressure sensor, combines the spatial coordinates output by the displacement detection unit, compares the movement direction with the pressure change trend, collects the dynamic response data of the acceleration detection unit, judges the pressure gradient and depth change at the contact point, and obtains the composite sensing state factor. Based on the composite sensing state factor, the state discrimination module determines the pressure gradient distribution of each contact area, analyzes the change characteristics of the continuous spatial area, compares the relationship between the end movement direction and the contact depth change direction, identifies the dynamic contact state, and obtains the contact interaction trend characteristics. The mode switching module adjusts the controller's response strategy to the interaction trend characteristics of the touch points, analyzes and identifies the state, changes the execution logic of the main control unit, filters the control channels that are compatible with the current massage action, adjusts the operation parameter allocation mode, uses the configuration parameter content as the switching basis, and obtains the action mode conversion signal. Based on the motion mode conversion signal, the instruction output module adjusts the controller proportional adjustment configuration, synchronously refreshes the motion execution rhythm of the end massage component, switches to the new motion state, and obtains a dynamic control instruction set. Based on the dynamic control instruction set, the calibration and adjustment module compares the contact feedback within the cycle, analyzes the gradient change of the composite sensing state factor in the previous cycle, determines the current deviation trend, adjusts the main control unit parameter settings and discrimination criteria, and obtains the synchronous deviation calibration result.

Images