Trajectory Generation Method, System and Medium Based on Action Intention Recognition

Abstract

The present invention relates to a trajectory generation method, system and medium based on action intention recognition. The trajectory generation method includes: obtaining the component of the force value exerted by a rehabilitation patient for the recovery position during the rehabilitation movement; identifying the movement intention of the rehabilitation patient based on the force value component; and generating the end trajectory of the rehabilitation exoskeleton robot according to the movement intention, where each position in the end trajectory is constrained within a preset spatial range. The trajectory generation method of the present invention collects the force exertion of the patient's wrist in real time through a force sensor, and accordingly accurately distinguishes and identifies the action intention, solves the problem of distinguishing the motion state of the end force sensor for the robotic arm, and facilitates the distinction between the start and stop states. It also quickly generates a suitable motion trajectory through position accumulation and spatial constraint to support the implementation of the motion, grasping and obstacle avoidance tasks of the robotic arm, and provides support for the motion tasks of the rehabilitation exoskeleton robot.

Description

Technical Field

[0001] The present invention relates to the technical field of trajectory generation, and in particular to motion trajectory generation of a mechanical arm suitable for a rehabilitation exoskeleton robot. A trajectory generation method, device, system and computer-readable medium based on action intention recognition are proposed. Background Art

[0002] Motion trajectory generation is a prerequisite for completing tasks such as robot arm movement, grasping, and obstacle avoidance. A good trajectory generation front end can provide good input for the accurate and timely execution of specific tasks at the back end and reduce the workload of back end optimization.

[0003] At present, the development of the classic collaborative robot arm SDK has been relatively complete. It can realize the mutual conversion of the collaborative arm end posture to each joint angle, that is, forward and inverse solution. For example, the posture angle of the collaborative arm end is expressed by Euler angle representation. When performing forward and inverse solution, the Euler angle is converted into a rotation matrix. When performing trajectory planning, the trajectory of the path point is generated through polynomial fitting function (such as fifth-order polynomial) and optimization solver.

[0004] However, the trajectory planning scheme of the existing collaborative robot is mainly designed to solve and plan the joint movement of the serial manipulator arm robot, and use the force sensor built into the joint motor to record the angular momentum of each joint during the dragging process in a dragging teaching manner. The optimization of the terminal motion trajectory of the rehabilitation exoskeleton robot is insufficient. The dragging method of the joint force sensor relied on by this type of trajectory planning cannot accurately evaluate the terminal input force of the exoskeleton robot after being worn (generated by the wearer under specific movement intentions), while patients are more concerned about the recovery level of the terminal force in the recovery position such as the hands. The existing trajectory planning method is not accurate enough in mapping the terminal force when the rehabilitation exoskeleton robot is used.

[0005] In addition, in the research on trajectory planning of collaborative robots, a robot-based modeling method was proposed, which combined with dynamic optimization algorithms, acceleration continuity and other methods to generate trajectory for the robot's end manipulator motion trajectory. However, this type of method also cannot focus on the mapping and evaluation of the end force in the recovery position such as the hands of rehabilitation patients. Summary of the invention

[0006] In view of the defects of the prior art, according to the first aspect of the present invention, a trajectory generation method based on motion intention recognition is proposed, which aims to achieve accurate recognition of motion intention and generate the terminal trajectory of the rehabilitation exoskeleton robot accordingly through real-time collection and analysis of the force conditions of the rehabilitation position of the rehabilitation patient.

[0007] As an optional example, a trajectory generation method based on action intention recognition includes the following steps:

[0008] Obtain the force value component of the recovered patient exerting force on the recovery position during the rehabilitation exercise;

[0009] Identify the motion intention of the recovered patient based on the force value component;

[0010] Generate the end trajectory of the rehabilitation exoskeleton robot according to the motion intention, and each position in the end trajectory is restricted within a preset space range.

[0011] As an optional example, the obtaining the force value component of the recovered patient exerting force on the recovery position during the rehabilitation exercise includes:

[0012] Obtain the force exerted by the recovered patient on the recovery position sensed by the three-dimensional force sensor in real time, and obtain the force value components of the X, Y, and Z axes; wherein, the three-dimensional force sensor is arranged at the recovery position of the recovered patient.

[0013] As an optional example, the identifying the motion intention of the recovered patient based on the force value component includes:

[0014] Obtain the statistical change trend of the force value component in each axis direction according to the force value component;

[0015] In response to the statistical change trend on any axis satisfying the preset condition, determine that there is a force input on the axis; and

[0016] Activate the motion intention flag bit.

[0017] Thus, based on the acquisition of the force actually exerted by the recovered patient on the recovery position, the motion intention is identified, and accordingly, the mapping of the actual force and intention of the patient is realized. The rehabilitation exoskeleton robot moves according to the motion trajectory generated according to this intention to assist the recovered patient in performing motion rehabilitation.

[0018] As an optional example, the obtaining the statistical change trend of the force value component in each axis direction according to the force value component includes:

[0019] According to the obtained force value components in the X, Y, and Z axis directions, respectively obtain the mean and variance of the force value components in each axis direction as the statistical change trend.

[0020] As an optional example, the in response to the statistical change trend on any axis satisfying the preset condition, determine that there is a force input on the axis includes:

[0021] If the mean and variance of the force value components on any axis both reach a preset steady state, it is determined that there is a force input on the axis. The satisfaction of the preset conditions includes that within a preset continuous sampling period, the change amount of the mean and the change amount of the variance are both within a preset threshold range, and it is determined that the preset conditions are satisfied.

[0022] Thus, by calculating the variance and mean in real time, the statistical change trend of the three force value components can be measured. The mean of the data represents the concentration level of the data, and the variance reflects its degree of dispersion. Therefore, when the mean is stable, the motion intention is recognized, it is determined that the rehabilitation patient hopes to use this force input to control the movement of the rehabilitation exoskeleton robot at this time, the motion intention flag bit is activated, and position accumulation and trajectory generation are performed accordingly. In other stages (i.e., unstable stages), it is not activated and does not participate in trajectory generation. Moreover, for each single-axis direction force input, the stable steady stage is taken as the activation stage of the motion intention flag bit to avoid possible jitter and ensure the smooth progress of the rehabilitation movement.

[0023] As an optional example, the generating the end trajectory of the rehabilitation exoskeleton robot according to the motion intention includes:

[0024] Based on the activated motion intention flag bit, motion position accumulation is performed using a preset search constraint rule, and the end trajectory of the rehabilitation exoskeleton robot is generated with the accumulated position sequence.

[0025] As an optional example, the motion position accumulation is set to be performed in the following manner:

[0026] For the force value component obtained on any axis, multiply the force value component at the current position by a preset step adjustment factor, and add the result to the accumulated force value component on the axis as the current accumulated position.

[0027] Thus, through the spatial search limitation under the preset constraint rule, it is ensured that the spatial search process will not exceed the limit range, and the search progress and efficiency can also be accelerated. The motion trajectory of the rehabilitation exoskeleton robot is generated through the trajectory sequence formed by position accumulation.

[0028] As an optional example, by adjusting the size of the preset step adjustment factor, motion trajectories corresponding to different ranges are generated to meet the requirements of different motion rehabilitation action tasks.

[0029] According to the second aspect of the object of the present invention, a computer system is further proposed, including:

[0030] One or more processors, and

[0031] A memory for storing operable instructions;

[0032] Wherein, when the instructions are executed by the one or more processors, the process of the aforementioned trajectory generation method based on action intention recognition is implemented.

[0033] According to a third aspect of the purpose of the present invention, there is also provided a computer-readable medium storing a computer program, the computer program includes instructions that can be executed by one or more processors, and when the instructions are executed by the one or more processors, the process of the aforementioned trajectory generation method based on action intention recognition is implemented.

[0034] Combined with the implementation of each of the above aspects, the trajectory generation method based on action intention recognition proposed by the present invention realizes efficient and accurate trajectory generation through the combination of motion intention recognition and differentiation, search space constraint, and trajectory integral cumulative motion position. Compared with the existing trajectory generation systems, the significant beneficial effects of the method proposed by the present invention are as follows:

[0035] (1) For the end trajectory of the rehabilitation exoskeleton robot, through the accurate differentiation of motion intention, the action intention of the user can be accurately recognized, whether to generate a trajectory can be selected, and a suitable motion trajectory can be generated to support the realization of tasks such as the movement, grasping, and obstacle avoidance of the robotic arm of the robot, solving the problem of differentiating the motion state of the robotic arm by the end force sensor, facilitating the differentiation of the start and stop states of the robot, and being able to avoid possible jitters, ensuring the smooth progress of the rehabilitation movement;

[0036] (2) It can, through the constraints within the trajectory sampling space, cut unnecessary sampling space as needed, generate safe and necessary trajectories, reduce the amount of calculation, and improve the efficiency of trajectory generation;

[0037] (3) It can generate trajectories with different ranges and different step sizes by flexibly adjusting the integral parameters to meet different motion requirements.

[0038] It should be understood that all combinations of the foregoing concepts and additional concepts described in more detail below can be regarded as part of the inventive subject matter of the present disclosure as long as such concepts do not contradict each other. In addition, all combinations of the claimed subject matter are regarded as part of the inventive subject matter of the present disclosure.

[0039] The foregoing and other aspects, embodiments, and features of the teachings of the present invention can be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the present invention, such as the features and / or beneficial effects of the exemplary embodiments, will be apparent from the following description or will be learned through the practice of the specific embodiments according to the teachings of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are not intended to be drawn to scale. In the accompanying drawings, each identical or nearly identical component shown in the various figures may be denoted by the same reference numeral. For the sake of clarity, not every component is labeled in each figure. Now, embodiments of various aspects of the present invention will be described by way of example and with reference to the accompanying drawings.

[0041] Figure 1 It is a schematic flowchart of a trajectory generation method based on action intention recognition according to an embodiment of the present invention.

[0042] Figure 2 It is a schematic flowchart of recognizing the motion intention of a rehabilitation patient based on the force value component according to an embodiment of the present invention.

[0043] Figure 3 It is a schematic diagram of a three-dimensional force sensor according to an embodiment of the present invention.

[0044] Figure 4 It is a schematic diagram of an execution process of a trajectory generation method based on action intention recognition according to an embodiment of the present invention.

[0045] Figure 5 It is a mean value curve graph of the force value component data based on a three-dimensional force sensor according to an embodiment of the present invention.

[0046] Figure 6 It is a variance curve graph of the force value component data based on a three-dimensional force sensor according to an embodiment of the present invention.

[0047] Figure 7 It is a motion intention recognition result graph of the force value component data based on a three-dimensional force sensor according to an embodiment of the present invention.

[0048] Figure 8 It is a schematic diagram of a spherical constraint sampling space according to an embodiment of the present invention.

[0049] Figure 9 It is a schematic diagram of a cube constraint sampling space according to an embodiment of the present invention.

[0050] Figure 10 It is a schematic diagram of a trajectory generated by the method according to an embodiment of the present invention under spherical constraints.

[0051] Figure 11 It is a schematic diagram of a trajectory generated by the method according to an embodiment of the present invention under cube constraints.

[0052] Figure 12 It is a schematic diagram of a trajectory generated by the method according to an embodiment of the present invention without constraints.

[0053] Figure 13 It is a test comparison graph of the method according to an embodiment of the present invention and the existing method. Detailed Implementation Modes

[0054] To better understand the technical content of the present invention, specific embodiments are hereby given and described in conjunction with the accompanying drawings as follows.

[0055] In the present disclosure, aspects of the present invention are described with reference to the accompanying drawings, in which many illustrative embodiments are shown. The embodiments of the present disclosure are not necessarily intended to cover all aspects of the present invention. It should be understood that the various concepts and embodiments introduced above, as well as those concepts and embodiments described in more detail below, can be implemented in any of many ways, because the concepts and embodiments disclosed in the present invention are not limited to any implementation manner. Additionally, some aspects of the present invention can be used alone, or in any suitable combination with other aspects of the present invention.

[0056] {Trajectory Generation Method Based on Action Intention Recognition}

[0057] According to the embodiments disclosed in the present invention, a trajectory generation method based on action intention recognition is proposed, which is particularly applicable to the motion trajectory generation of the robotic arm of a rehabilitation exoskeleton robot. It aims to collect the force exerted on a rehabilitation patient by a three-dimensional force sensor and recognize the action intention, and on this basis, achieve an efficient and fast trajectory generation method, providing good support for the motion tasks of the rehabilitation exoskeleton robot, meeting the motion trajectory planning requirements of different types of rehabilitation exoskeleton robots, and realizing the function of "starting to move as soon as hands touch, stopping as soon as hands leave".

[0058] As an example, as Figure 1 shown in the exemplary embodiment of the trajectory generation method based on action intention recognition, it includes the following steps:

[0059] Step S101: Obtain the component of the force value when a rehabilitation patient exerts force towards the recovery position during the rehabilitation movement;

[0060] Step S102: Based on the force value component, recognize the motion intention of the rehabilitation patient;

[0061] Step S103: According to the motion intention, generate the end trajectory of the rehabilitation exoskeleton robot, and each position in the end trajectory is restricted within a preset spatial range.

[0062] As an optional embodiment, in the aforementioned step S101, obtaining the component of the force value when a rehabilitation patient exerts force towards the recovery position during the rehabilitation movement includes:

[0063] Obtain the force exerted by the three-dimensional force sensor on the recovery position of the rehabilitation patient in real time to obtain the force value components of the X, Y, and Z axes; wherein, the three-dimensional force sensor is arranged at the recovery position of the rehabilitation patient.

[0064] It should be understood that, in combination with Figure 3 the example shown, the aforementioned three-dimensional force sensor is configured as a sensor device capable of performing real-time decomposed force detection in three directions of X, Y, and Z. The three-dimensional force sensor is installed at a predetermined position of the rehabilitation patient in an appropriate form, such as joint positions like the wrist and ankle. After calibration and initialization, it can continuously and real-time collect the force exerted by the rehabilitation patient on the rehabilitation position (i.e., representing the motion rehabilitation training intention of the rehabilitation patient), and obtain the decomposed amount data of the force values in the X, Y, and Z axis directions, that is, the force value components on each axis.

[0065] For the sake of convenience in description, in the following embodiments, the rehabilitation training of the wrist is taken as an example for illustration. Through the three-dimensional force sensor, by collecting the force exertion situation of the patient's wrist, the components on the X, Y, and Z axes are obtained, accurately identifying the user's motion intention, solving the problem of distinguishing the motion states of the end-effector force sensor for the robotic arm, facilitating the rehabilitation exoskeleton robot to distinguish the start and stop states, and enabling the distinction of motion intentions at any hovering angle, and can conveniently implement the function of "starting to move as soon as touching, stopping as soon as leaving".

[0066] In the embodiments of the present invention, the resultant force vector of the three-dimensional force sensor is represented by the components on the X, Y, and Z axes, and the components of the force values on each axis (unit: N) are represented in the form of a matrix, that is:

[0067]

[0068] wherein, , , respectively represent the force value components on the X, Y, and Z axes.

[0069] In combination with Figure 2 the example shown, in the aforementioned step S102, based on the force value components, identifying the motion intention of the rehabilitation patient includes:

[0070] Step S201: According to the force value components, obtain the statistical change trend of the force value components in each axis direction;

[0071] Step S202: In response to the statistical change trend on any axis satisfying a preset condition, determine that there is a force input on the axis;

[0072] Step S203: Activate the motion intention flag bit.

[0073] In the embodiments of the present invention, the mean value and variance of the force value components are used to judge and distinguish the motion intention of the rehabilitation patient.

[0074] As an alternative embodiment, obtaining the statistical change trend of the force value component in each axis direction according to the force value component includes:

[0075] According to the obtained force value components in the X, Y, and Z axis directions, the mean and variance of the force value component in each axis direction are respectively obtained as the criteria for the statistical change trend.

[0076] As an alternative embodiment, the determining that there is a force input on an axis in response to the statistical change trend on any axis satisfying a preset condition includes:

[0077] In response to the mean and variance of the force value component on any axis both reaching a preset stable state, it is determined that there is a force input on the axis.

[0078] As an alternative example, the foregoing satisfaction of the preset condition includes:

[0079] Within a preset continuous sampling period, if the change amount of the mean and the change amount of the variance are both within a preset threshold range, it is determined that the preset condition is satisfied.

[0080] Thus, through the variance and the mean, the statistical change trend of the three force value components can be measured, where the mean of the data represents the concentration level of the data, and the variance reflects its degree of dispersion.

[0081] On this basis, by continuously obtaining and storing the force value component data of the three axes, and calculating the mean and variance of the force value component data obtained on the three axes in real time, when both the mean and the variance are in a relatively stable state, it is determined that the rehabilitation patient hopes to use this force input to control the movement of the robot for motion rehabilitation training. Accordingly, the motion intention flag bit is activated, and the motion intention flag bit is not activated in other stages.

[0082] Combined with Figure 1 As shown, as an alternative embodiment, the foregoing generating the end trajectory of the rehabilitation exoskeleton robot according to the motion intention includes:

[0083] Based on activating the motion intention flag bit, the motion positions are cumulatively calculated using a preset search constraint rule, and the end trajectory of the rehabilitation exoskeleton robot is generated with the cumulative position sequence.

[0084] In the embodiment of the present invention, only when the motion intention flag is activated can the motion positions be cumulatively calculated to form a position sequence for guiding the movement of the robot to assist the rehabilitation patient for rehabilitation positions, such as wrist rehabilitation training.

[0085] As an alternative example, the foregoing cumulative calculation of the motion positions is set to be performed in the following manner:

[0086] For the force value component obtained on any axis, multiply the force value component at the current position by a preset step adjustment factor, and add the result to the accumulated force value component on the axis as the current accumulated position.

[0087] In an embodiment of the present invention, an overly large search space may affect the calculation speed of the integrator. Additionally, in some scenarios, there are requirements for the safety range of the trajectory. Therefore, it is necessary to limit the range of the search space. As an example, the space search of the present invention uses the following three forms of constraint rules to limit the trajectory space, namely: unconstrained form, spherical constraint form, and cubic constraint form.

[0088] Such as Figure 8 , 9 shown respectively represent examples of the sampling spaces of the spherical constraint form and the cubic constraint form.

[0089] Thus, in the process of generating a trajectory in an embodiment of the present invention, based on the constraint method within the trajectory sampling space, unnecessary sampling spaces can be trimmed as needed to generate safe and necessary trajectories, and the efficiency of space search and trajectory generation can be improved, enabling the rapid generation of suitable motion trajectories to support the execution of various rehabilitation training action tasks of the robotic arm of a robot, including but not limited to: swinging, grasping, and obstacle avoidance.

[0090] In an optional embodiment, the aforementioned step adjustment factor serves as an integration parameter and can be preset and adjusted in size to adjust the size of a single step, thereby achieving the generation of trajectories in different large, medium, and small ranges.

[0091] In some embodiments, by adjusting the aforementioned integration parameter, i.e., the step adjustment factor, trajectories in different ranges can be generated to meet the requirements of different rehabilitation movement actions. Therefore, in the trajectory generation method proposed by the present invention, by flexibly adjusting the integration parameter, more flexible and diverse trajectory generation can be achieved to adapt to the requirements of different rehabilitation training movement tasks.

[0092] {Embodiment 1}

[0093] To more specifically elaborate on the implementation of the trajectory generation method of the present invention, we will further describe an exemplary implementation of the aforementioned trajectory generation method in combination with the attached Figure 3 - 12 Take the wrist movement recovery of the upper limb of a rehabilitation patient as an example and describe a specific process of the trajectory generation process shown in combination with Figure 4 for illustration.

[0094] The implementation process of the trajectory generation method in this embodiment is configured to be implemented through four parts, namely, obtaining the force value component of the rehabilitation patient, judging the motion intention, constraining the search space, and generating a trajectory based on the space constraint. We will describe in combination with Figure 4And Figure 5 - 12 The implementation of the above four parts is described in more detail as shown below.

[0095] 1.1 Wrist Force Analysis and Acquisition of Triaxial Force Value Components

[0096] In this example, by deploying a three-dimensional force sensor on the wrist of a rehabilitation patient, the force application situation of the patient's wrist is collected according to a preset sampling period, and the decomposed force value, that is, the component force data, is output.

[0097] The following matrix represents the matrix expression of a resultant force vector obtained by collecting through a three-dimensional force sensor:

[0098]

[0099] That is, it means that the force value components on the X, Y, and Z axes are 3.12N, -0.4N, and 5.12N respectively.

[0100] On this basis, the modulus of the resultant force vector is obtained, which is expressed as:

[0101] .

[0102] Thus, the angles between the resultant force vector and the X, Y, and Z axes , , (in radians) are obtained and are respectively expressed as:

[0103]

[0104] Thus, based on the aforementioned angles , , , the direction and quadrant of the resultant force vector can be determined, and the magnitude of the resultant force can be determined according to the modulus of the resultant force vector.

[0105] In an alternative embodiment, the force component acquisition module can be configured as a data interface to continuously obtain the component data collected by the sensor.

[0106] 1.2 Motion Intention Judgment

[0107] In this example, based on the mean and variance of the force value components on the X, Y, and Z axes obtained by continuous acquisition, the motion intention is judged and distinguished.

[0108] As an alternative example, the mean calculation method of the sampled data on the three axes is as follows:

[0109]

[0110] Among them, , , respectively represent the mean values of the force value components along the X, Y, and Z axes, n represents the number of sampling times of the three-dimensional force sensor, that is, the number of samples of the component data on each axis.

[0111] , , respectively represent the i th sampling values of the force value components along the X, Y, and Z axes, that is, the observed values.

[0112] As an optional example, the variance calculation method for the sampling data on the three axes is as follows:

[0113]

[0114] wherein, , , respectively represent the variances of the force value components along the X, Y, and Z axes.

[0115] Thus, by calculating the variance and the mean value, the statistical change trend of the three-axis component data can be obtained. The mean value of the data represents the concentration level of the data, and the variance reflects its degree of dispersion.

[0116] In this example, by continuously obtaining the collected three-axis component data and calculating the mean value and variance on the three axes in real time, when the mean value and variance are stable, for example, their change amounts or change rates are within the allowable threshold range, it is determined that the rehabilitation patient hopes to use this force input to control the movement of the robotic arm at this time, and the motion intention flag bit is activated, and it is not activated in other stages.

[0117] As an optional method, the above real-time calculation of the mean value and variance and the motion intention judgment are implemented by configuring a motion intention judgment module. For example, the motion intention judgment module realizes the motion intention judgment and differentiation by maintaining the data containers of the force value components on the three axes and the motion intention flag bit, and outputs the judgment result.

[0118] Among them, the data containers of the force value components on the three axes are set to save the adopted data of the three-dimensional force sensor in a matrix data structure in real time, and calculate the mean value and variance of the three-axis component data in the container in real time. When both the mean value and variance are stable, the motion intention flag bit is activated and output, and participates in the trajectory generation, and it is not activated in other stages (that is, does not participate in the trajectory generation).

[0119] Combined with Figure 5 , 6 As shown in Fig. 7, taking the X axis as an example to illustrate the motion intention judgment process.

[0120] When the mean value of the X-axis sampling data of the three-dimensional force sensor exceeds a preset threshold, it can be considered that a certain force input has occurred in the axial direction of the sensor. Figure 5 Exemplarily shows the process of the mean value rising, stabilizing, and then releasing under the force input. The mean value will experience four processes: a sudden increase in stages (stages 10 to 12, stages 50 to 54), fluctuations (stages 12 to 20, stages 52 to 62), stability (stages 20 to 30, stages 60 to 70), and release to zero (stages 30 to 50, stages after 70).

[0121] Combined with Figure 6 As shown, compared with the change process of the mean value, the variance also experiences four processes: first rising suddenly, fluctuating, then stabilizing within a certain range, decreasing, and then stabilizing within a range, and finally returning to zero. As Figure 6 Exemplarily shows the data change during the same period.

[0122] Combined with Figure 7 As shown, for each single-axis force input process, in the embodiments of the present invention, only the relatively stable stable stage is taken as the motion intention flag bit as the activation stage to avoid possible partial jitter, such as Figure 7 Exemplarily shows the data change during the same period.

[0123] Combined with Figure 5 、 6 、7 shown, in the embodiments of the present invention, the motion intention is distinguished by using the variance and the mean value for synchronous comparison. By analyzing the variance value characteristics and mean value characteristics of the user, the motion intention can be distinguished according to the stage calculation results to accurately identify the user's action intention and provide accurate input and stop states for subsequent trajectory generation.

[0124] 1.3 Search space constraint

[0125] Too large a search space may affect the subsequent calculation speed of using the integrator for position accumulation. At the same time, due to the safety range limit requirements for the trajectory in the training and rehabilitation scenario, in the embodiments of the present invention, the range of the search space can be restricted according to preset rules.

[0126] As an optional embodiment, in the examples of the present invention, the following three forms of constraint methods are used to limit the trajectory space, specifically including: spherical constraint form, cubic constraint form, and unconstrained form.

[0127] 1.3.1 Spherical constraint

[0128] First, the cumulative values on the three axes ( 、 、 ) By cumulatively updating the input values of the force value components of each axis at the current moment, the sphere radius radius has been pre-configured and specified.

[0129] The min and max functions are used in the position cumulative update process to ensure that the cumulative value is within the range of - radius to radius , that is:

[0130] .

[0131] Then, by calculating the three-axis Euclidean distance distance to measure the length of the cumulative vector.

[0132] As an example, the aforementioned Euclidean distance is calculated using the square root of the sum of the squares of the values on the three coordinate axes, that is:

[0133]

[0134] If the distance exceeds radius , a scaling operation is required to ensure that the exceeded part is on the limit boundary and the non-exceeded part is within the limit boundary.

[0135] Among them, the scaling factor Scale_factor is calculated as: radius divided by distance . Then, the cumulative value is scaled by multiplying it by the scaling factor to ensure that the length of the cumulative vector does not exceed the sphere radius radius .

[0136] Thus, through the scaling operation, all points exceeding the constraint range are restricted to the boundary of the constraint range, and the previous search direction is maintained.

[0137] Such as Figure 8 exemplarily shows the distribution of each random point in the space after using the random dotting method with a normal distribution to dot the space and performing constraint scaling on the final result. It can be seen that all randomly distributed points are constrained inside the sphere, and the space constraint is successful.

[0138] As an optional example, the process of spherical constraint includes:

[0139] 1) Obtain the cumulative values on the three axes , , :

[0140] 2) Obtain the three-axis Euclidean distance distance ;

[0141] 3) Judge the three-axis Euclidean distance distanceIs it greater than the sphere radius? radius If it is greater than the sphere radius, calculate the scaling factor Scale_factor and based on the scaling factor Scale_factor perform a scaling operation on the accumulated values of the three axes to ensure that the length of the accumulated vector does not exceed the sphere radius radius The accumulated position is constrained within the sphere space.

[0142] 1.3.2 Cubic space constraint

[0143] Based on the above spherical constraint, for the cubic constraint, there is no need for the scaling operation of the sphere constraint. Only the first constraint operation of the sphere constraint needs to be retained to limit it within the corresponding range.

[0144] As an alternative embodiment, the min and max functions are used during the position accumulation update process to ensure that the accumulated value is within the range of - length to length , that is:

[0145]

[0146] where length represents the semi-axis length of the cube.

[0147] Figure 9 Exemplarily shows the distribution of randomly placed points in the space using the random dotting method with a normal distribution and the randomly placed points inside the space after constraint scaling of the final result. It can be seen that all randomly distributed points are constrained within the cube range, and the space constraint is successful.

[0148] 1.3.3 Unconstrained form

[0149] The unconstrained search space is a simple update operation on the accumulated value. The accumulated value on each axis is updated by adding the force input values of each axis at the current moment.

[0150] The position update method using the unconstrained form directly adds the coordinate value of the current point to the accumulated value without range limitation, which means that the accumulated value can exceed the cube range or the sphere range.

[0151] In the embodiments of the present invention, the operation can be accelerated by considering the influence of different search spaces. For example, by selecting the spherical constraint method or the cubic constraint method, the calculation amount can be reduced and the efficiency of trajectory generation can be improved by constraining the search space. By reasonably designing the constraint conditions of the search space, the operation speed can be increased on the premise of ensuring accuracy.

[0152] As an alternative, the spatial constraint of the predefined rules is implemented by configuring the search space constraint module, including but not limited to spherical constraint, cube constraint, or unconstrained rules.

[0153] 1.4 Trajectory Generation

[0154] In this example, the generation of the motion trajectory of the rehabilitation robot manipulator is achieved by configuring a trajectory generator. Among them, whether the trajectory generator is started is executed according to the result of the aforementioned motion intention judgment module, that is, if there is a motion intention, the generation is started, otherwise the generation is not started.

[0155] In this example, when the trajectory generator responds to the generation of the motion intention (i.e., the motion intention flag bit is activated), it calls the search space constraint module to limit the search space according to the preset rules to ensure that the spatial search process does not exceed the limit range and can also accelerate the search.

[0156] After the search space constraint module completes the spatial constraint, the specific trajectory generation work is completed by a position accumulator, which is used to calculate the cumulative position and perform limitation and scaling on it.

[0157] The position accumulator is internally configured with a position sequence container for maintaining and real-time updating the cumulative position sequences of the three axes.

[0158] Taking the X-axis as an example, in one calculation, the X-axis force value component of the current position is multiplied by the step size adjustment factor and added to the cumulative X-axis component. Then, it is checked whether the three-axis Euclidean distance of the cumulative position exceeds the limit range. If it exceeds the limit range, corresponding scaling and constraint are performed.

[0159] In this example, taking the spherical constraint as an example, the sphere radius is preconfigured, and by calculating the scaling factor, the cumulative position is scaled to the limit range of the spherical space.

[0160] At the same time, each component of the cumulative position is multiplied by the scaling factor to achieve overall scaling.

[0161] In the above manner, through the processing of the force value components on the continuously obtained three axes, the integral result of the position accumulator will not exceed the scaled limit position at most.

[0162] As Figure 10 shown, an example of the trajectory generated using the spherical constraint rule is exemplarily represented.

[0163] In another example, we adopt the cube constraint and the unconstrained rule to generate trajectories in combination with the motion intention, and the formed trajectories are respectively as Figure 11 and 12 shown.

[0164] In an alternative embodiment, the aforementioned step size adjustment factor serves as an integral parameter, and its magnitude can be preset and adjusted to adjust the size of the single-step size, thereby achieving the generation of trajectories in different large, medium, and small ranges.

[0165] In some embodiments, by adjusting the aforementioned integral parameter, i.e., the step size adjustment factor, trajectories in different ranges can be generated to meet the requirements of different rehabilitation movement actions. Therefore, in the trajectory generation method proposed by the present invention, by flexibly adjusting the integral parameter, more flexible and diverse trajectory generation can be achieved to adapt to the requirements of different rehabilitation training movement tasks.

[0166] 1.5 Test Data Comparison

[0167] Combined with Figure 13 As shown, we will compare the results obtained by using the trajectory generation method based on action intention of the above embodiments of the present invention with the trajectory generation method based on minimaljerk dynamics optimization in the prior art. 70 data points are generated respectively by the two methods. The trajectory generation method based on minimaljerk uses a fifth-degree polynomial to calculate the parameter matrix of 70 path points, and the method of the above embodiments of the present invention is used to calculate the generated position points of 70 path points. By recording the time consumption of 5 groups of data, combined with Figure 13 It can be seen that the time consumed for path calculation by the method of the present invention has been significantly improved, and the time consumption is shortened by more than 95%.

[0168] {Embodiment 2}

[0169] Combined with the implementation of the trajectory generation method based on action intention recognition in the above embodiments, according to the embodiments disclosed in the present invention, a computer system is further proposed, including: one or more processors, and a memory for storing operable instructions.

[0170] Wherein, when the instructions are executed by the one or more processors, the process of the trajectory generation method based on action intention recognition in any of the foregoing embodiments is implemented.

[0171] {Embodiment 3}

[0172] Combined with the implementation of the trajectory generation method based on action intention recognition in the above embodiments, according to the embodiments disclosed in the present invention, a computer-readable medium storing a computer program is further proposed. The computer program includes instructions executable by one or more processors, and when the instructions are executed by the one or more processors, the process of the trajectory generation method based on action intention recognition in any of the foregoing embodiments is implemented.

[0173] As an alternative example, the foregoing computer-readable medium may be implemented by including, but not limited to, random access memory, read-only memory, electrically erasable programmable memory, optical disk memory, magnetic disk memory, and combinations of the foregoing types of computer-readable storage media.

[0174] Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Those with ordinary knowledge in the technical field to which the present invention pertains may make various modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to what is defined by the claims.

Claims

1. A trajectory generation method based on action intention recognition, characterized in that Including the following steps: Obtain the component of the force value when the rehabilitated patient exerts force on the recovery position during the rehabilitation exercise; Based on the force value component, identify the motion intention of the rehabilitated patient, specifically including: according to the force value component, obtain the statistical change trend of the force value components in the X, Y, and Z axis directions; in response to the statistical change trend on any axis satisfying the preset condition, determine that there is a force input on the axis, that is, determine that the rehabilitated patient hopes to use this force input to control the movement of the robot at this time, and accordingly activate the motion intention flag bit; Generate the end trajectory of the rehabilitation exoskeleton robot according to the motion intention, specifically including: based on the activated motion intention flag bit, perform cumulative motion positions using a preset search constraint rule, and generate the end trajectory of the rehabilitation exoskeleton robot with the cumulative position sequence; each position in the end trajectory is constrained within a preset space range.

2. The trajectory generation method based on action intention recognition according to claim 1, wherein The obtaining of the component of the force value when the rehabilitated patient exerts force on the recovery position during the rehabilitation exercise includes: Obtain the force exerted by the rehabilitated patient on the recovery position in real time by a three-dimensional force sensor, and obtain the force value components of the X, Y, and Z axes; Wherein, the three-dimensional force sensor is arranged at the recovery position of the rehabilitated patient.

3. The trajectory generation method based on action intention recognition according to claim 1, characterized in that The obtaining of the statistical change trend of the force value component in each axis direction according to the force value component includes: According to the obtained force value components in the three directions of the X, Y, and Z axes, respectively obtain the mean value and variance of the force value component in each axis direction as the statistical change trend.

4. The trajectory generation method based on action intention recognition according to claim 3, wherein The response to the statistical change trend on any axis satisfying the preset condition and determining that there is a force input on the axis includes: In response to the mean value and variance of the force value component on any axis both reaching the preset stable state, determine that there is a force input on the axis.

5. The trajectory generation method based on action intention recognition according to claim 3, characterized in that The satisfaction of the preset condition includes: Within a preset continuous sampling period, the change amount of the mean value and the change amount of the variance are both within the preset threshold range, and it is determined that the preset condition is satisfied.

6. The trajectory generation method based on action intention recognition according to claim 1, wherein The cumulative motion position is set to be performed in the following manner: For the force value component obtained on any axis, multiply the force value component of the current position by a preset step adjustment factor, and add the result to the cumulative force value component on the axis as the current cumulative position.

7. The trajectory generation method based on action intention recognition according to claim 6, wherein The method further includes the following steps: Adjust the size of the preset step adjustment factor to generate motion trajectories corresponding to different ranges.

8. A computer system, characterized in that, Including: One or more processors, and a memory for storing instructions that can be operated; Wherein, when the instructions are executed by the one or more processors, the process of the trajectory generation method based on action intention recognition according to any one of claims 1-7 is implemented.

9. A computer-readable medium storing a computer program, characterized in that, The computer program includes instructions that can be executed by one or more processors, and when the instructions are executed by the one or more processors, the process of the trajectory generation method based on action intention recognition according to any one of claims 1-7 is implemented.

Images