Method for generating anthropomorphic walking gait of self-balancing lower limb exoskeleton

Abstract

The invention discloses an anthropomorphic walking gait generation method of a self-balancing lower limb exoskeleton. The method comprises the following steps: initializing a reference zero moment point track, a centroid sagittal plane track and a sole track according to a human body walking motion rule; generating transverse movement of the self-balancing lower limb rehabilitation exoskeleton by adopting a previewing control method based on the initialized reference zero moment point track; on the basis of the initialized centroid sagittal plane track and the sole track, sagittal motion is generated through a straight knee optimizer and a stable filter, the straight knee optimizer is used for optimizing hip joint tracks of different heights, and the stable filter is used for generating hip joint tracks meeting dynamic constraints; and solving the inverse kinematics by adopting a Levenberg-Marquardt algorithm to obtain each joint track of the self-balancing lower limb rehabilitation exoskeleton. By utilizing the method, the anthropomorphic walking gait with the characteristics of mass center height change, knee straightening, heel landing and toe off-ground can be generated.

Description

Technical Field

[0001] The present invention relates to the technical field of robots, and more particularly, to a method for generating an anthropomorphic walking gait of a self-balancing lower-limb exoskeleton. Background Art

[0002] A self-balancing lower-limb rehabilitation exoskeleton (SLLRE) is a type of exoskeleton robot with high degrees of freedom, full drive, and the ability to walk dynamically balanced without additional auxiliary support tools such as crutches or carts. It can assist hemiplegic, paraplegic, and quadriplegic patients in performing lower-limb medical rehabilitation training, provide support instead of a wheelchair, and achieve dynamically balanced walking. In physical therapy, the goal of gait rehabilitation training is to help people with lower-limb dysfunction recover the gait of a healthy person. Therefore, generating the gait of a healthy person for the SLLRE is a key factor in improving the wearing comfort and rehabilitation energy efficiency of the SLLRE.

[0003] Currently, gait planning methods can be divided into model-based methods and model-free methods. Model-based methods first require establishing a physical model of the robot. Since the SLLRE has a high degree of freedom and is a very complex dynamic system with multivariable non-linear characteristics, it is difficult to establish an accurate dynamic model. Therefore, it is usually modeled as a common simple physical model such as an inverted pendulum model, a cart-table, a linkage model, etc. Then, the gait is planned according to the model. Taking the inverted pendulum model as an example, the SLLRE can be approximated as a linear inverted pendulum composed of a mass point that concentrates all the mass and a massless leg connecting the mass point to the support point. The moment at the support point is zero, so it is also called the Zero Moment Point (ZMP). Then, the walking cycle, walking parameters (such as step length, step width, single- and double-support times, etc.), the initial position of the center of mass, and the initial landing point (x0, y0) are set for the model, and then the expected landing point (x1, y1) of the next cycle is calculated according to the corresponding dynamic model, and the recursion is carried out in turn until the final landing point (X n , y n ) is obtained. After obtaining the landing point trajectory, the ZMP trajectory also needs to be planned, and then the corresponding center of mass trajectory is calculated using the linear inverted pendulum equation. Finally, the trajectories of each joint are obtained through inverse kinematics from the ZMP trajectory and the center of mass trajectory. This method can meet the stability of the walking process, but since the solution is not unique and there are optimization problems, it cannot guarantee the anthropomorphic degree of the gait.

[0004] Model-free gait generation methods usually constrain and control bipedal walking motion from two aspects: the stability and energy of the robot. By using stability as a constraint condition and minimizing energy consumption as an optimization goal to plan the gait of the SLLRE, its walking pattern can be made closer to that of humans. This method can give full play to the performance of the robot and reduce walking energy consumption. However, the optimal planning has a large amount of calculation, and there may be no solution in complex situations.

[0005] In the prior art, in the field of self-balancing lower limb exoskeletons, there is no technical solution for anthropomorphic gait generation. In the field of humanoid robots, a humanoid robot gait generation method has been proposed. The steps of this method include: establishing an initial walking pattern generator by analyzing the driving characteristics of humans to generate joint driving torques that can be used to achieve basic walking motion; establishing a stretch reflex mathematical model and a mathematical model of the vestibular reflex through the local characteristics of sensory reflexes, and superimposing the joint driving torques generated by the stretch reflex and the joint driving torques generated by the vestibular reflex on the joint driving torques; adjusting the joint driving torques to adjust the joint trajectories of the robot in the motion space to ensure that the supporting leg does not leave the ground and the swinging leg can reach the correct position during the walking process of the robot; establishing a ZMP reflex model, superimposing the joint driving torques generated by the ZMP reflex on the joint driving torques, and further adjusting the above joint trajectories by adjusting the joint driving torques to enable the robot to achieve stable walking motion and improve the environmental adaptability of the robot's motion.

[0006] After analysis, although the current model-based gait generation methods can generate stable walking gaits well, they cannot guarantee the anthropomorphic degree of the gaits. While the model-free gait generation methods can guarantee the anthropomorphic degree of the gaits, they have a large amount of calculation and cannot guarantee the feasibility of the solutions. Summary of the Invention

[0007] The object of the present invention is to overcome the defects of the above-mentioned prior art and provide an anthropomorphic walking gait generation method for self-balancing lower limb exoskeletons. This method includes the following steps:

[0008] Initialize the reference zero moment point trajectory, the center of mass sagittal plane trajectory, and the sole trajectory according to the human walking motion law;

[0009] Based on the initialized reference zero moment point trajectory, use preview control method to generate the lateral motion of the self-balancing lower limb rehabilitation exoskeleton;

[0010] Based on the initialized center of mass sagittal plane trajectory and the sole trajectory, use a straight-knee optimizer and a stability filter to generate sagittal motion, where the straight-knee optimizer is used to optimize the hip joint trajectories at different heights, and the stability filter is used to generate hip joint trajectories that meet the dynamic constraints;

[0011] The Levenberg-Marquardt algorithm is used to solve the inverse kinematics, and the joint trajectories of the self-balancing lower limb rehabilitation exoskeleton are obtained.

[0012] Compared with the prior art, the advantages of the present invention are as follows. For the series-parallel hybrid complex mechanism with multiple degrees of freedom and multiple rigid bodies of the self-balancing lower limb exoskeleton robot, in order to improve the wearing comfort and rehabilitation energy efficiency of the SLLRE, the three-dimensional motion of the SLLRE is decoupled based on the inverted pendulum model in the present invention. Adopting the idea of hierarchical optimization, the motion in each single-dimensional direction is optimized sequentially. When optimizing, the human walking motion law is satisfied as much as possible. Finally, the motions in each direction are coupled into the three-dimensional anthropomorphic walking gait of the SLLRE, and a more suitable anthropomorphic gait is designed for the SLLRE.

[0013] Other features and advantages of the present invention will become clear from the following detailed description of the exemplary embodiments of the present invention with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The drawings incorporated in and constituting a part of this specification illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0015] Figure 1 is a flowchart of a method for generating an anthropomorphic walking gait of a self-balancing lower limb exoskeleton according to an embodiment of the present invention;

[0016] Figure 2 is a schematic diagram of an anthropomorphic walking gait generation framework according to an embodiment of the present invention;

[0017] Figure 3 is a schematic diagram of initializing the sole trajectory based on a finite state machine according to an embodiment of the present invention;

[0018] Figure 4 is a schematic diagram of the virtual leg length constraint in a knee optimizer according to an embodiment of the present invention;

[0019] Figure 5 is a graph of the hip joint position and a schematic diagram of the virtual lengths of the left and right legs obtained by a straight knee optimizer according to an embodiment of the present invention;

[0020] Figure 6 is a schematic diagram of the hip joint position curve and the velocity curve after being processed by a stability filter according to an embodiment of the present invention;

[0021] Figure 7 is a schematic diagram of an anthropomorphic walking gait simulation experiment of a self-balancing lower limb exoskeleton according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0022] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present invention.

[0023] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way a limitation on the present invention or its application or use.

[0024] Techniques, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the techniques, methods, and devices should be regarded as part of the specification.

[0025] In all examples shown and discussed herein, any specific values should be construed as merely exemplary and not as a limitation. Thus, other examples of exemplary embodiments may have different values.

[0026] It should be noted that: like reference numerals and letters denote like items in the following drawings, and thus, once an item is defined in one drawing, further discussion thereof is not required in subsequent drawings.

[0027] Generally speaking, the anthropomorphic walking gait generation method for the self-balancing lower limb exoskeleton provided by the present invention includes: analyzing the human walking motion law, extracting walking characteristic parameters such as the height of the center of mass change, the toe-off angle, the knee joint extension angle, etc.; initializing the reference ZMP trajectory, and using preview control algorithm to generate the lateral motion of the SLLRE; initializing the center of mass sagittal plane trajectory and the sole trajectory, and using a straight knee optimizer and a stability filter to ensure that the knee joint is as straight as possible during the single-support phase; using the Levenberg-Marquardt algorithm (LM) to solve the inverse kinematics to obtain the trajectories of each joint of the robot.

[0028] Specifically, as shown in combination with Figure 1 and Figure 2 the anthropomorphic walking gait generation method for the self-balancing lower limb exoskeleton provided specifically includes the following steps:

[0029] Step S1, according to the human walking motion law, initialize the reference zero moment point trajectory, the center of mass sagittal plane trajectory, and the sole trajectory.

[0030] For example, the reference ZMP (zero moment point) and the sagittal plane center of mass trajectory are mainly determined according to the number of steps, step length, and period, and the sole trajectory is determined based on a finite state machine and quintic polynomial interpolation, and the state phases such as Figure 3 shown include the toe-off phase, the single-support phase, and the heel-strike phase. In the sagittal plane, the position of the sole can be determined by (x, z, θ), and the specific calculation method of each parameter is shown in formula (1).

[0031]

[0032] Among them, and are the position of the ankle joint of the swinging leg in the x-axis direction of the coordinate axis, the position in the z-axis direction of the coordinate axis, and the angle on the pitch joint, respectively. L stride is the step length, L forefoot and L heel represent the length of the forefoot and the length of the heel, respectively. θ toe and θ heel represent the pitch angle limits of the toes and the heels, respectively. H ankle represents the height of the ankle joint, H lift represents the maximum height of the foot sole lift. T c , T to , T hs and T m represent the single gait cycle, the toe-off phase cycle, the heel-strike phase cycle, and the single-support phase cycle, respectively. k represents a natural number (taking values 0, 1, 2, 3…, n), and t represents the moment. Figure 2 Among them, P x is the position of the ZMP on the x-axis, P y is the position of the ZMP on the y-axis. represents the reference position of the center of mass in the x-axis direction, y CoM represents the position of the center of mass in the y-axis direction, z CoM represents the position of the center of mass in the z-axis direction, x CoM is the position of the ankle joint on the x-axis, z ankle is the height of the ankle joint on the z-axis, θ pitch represents the pitch angle.

[0033] The sagittal plane trajectory of the center of mass is determined according to the designed number of steps, step length, and period to obtain the start and end positions of the hip joint in the x direction, and is obtained by piecewise interpolation using a cubic polynomial.

[0034] Step S2, based on the center of mass dynamics model, uses the preview control algorithm to generate the lateral movement of the center of mass.

[0035] Specifically, according to the center of mass dynamics model, the dynamic equation of the self-balancing lower limb exoskeleton in the coronal plane can be expressed as:

[0036]

[0037] Among them, y and represent the position and acceleration of the center of mass, respectively. p y represents the position of the ZMP on the y-axis, z cis the height of the centroid in the world coordinate system, and g is the acceleration due to gravity. Since the centroid dynamics equation (2) is expressed in continuous space, in practical applications, it needs to be rewritten as an expression in discrete space. Assuming the sampling time is Δt, the jerk of the centroid is used as the control input u(i) of the system, that is Then the discrete state equation is:

[0038]

[0039] where, A is the system matrix, representing the dynamic relationship between the internal state variables of the system, B is the input matrix, representing how the input signal affects the state variables of the system, and C is the output matrix, representing the relationship between the output of the i system and the state variables. Their values are respectively The current input of the system is The output of the system at the next moment is

[0040] To solve the ZMP deviation generated by the long-distance walking mode, the state space equation is rewritten in the following extended form:

[0041]

[0042] where, the extended control input is Δu(i) = u(i) - u(i - 1), the intermediate variable ΔΦ(i) = Φ(i) - Φ(i - 1), and the extended system input is The extended system matrix The extended input matrix The extended output matrix

[0043] where, Φ(i + 1) represents the output at the next moment, p y (i) represents the position of the ZMP in the y-axis direction at the i-th moment, and Φ * (i + 1) represents the extended output at the next moment.

[0044] To better track the ZMP, the walking problem can be written as a linear quadratic optimal control problem, and the quadratic performance index can be expressed as:

[0045]

[0046] where Q and R are positive weighting factors, represents the position of the reference ZMP in the y-axis at the j-th moment, J represents the performance index, also called the cost function, j is the time variable, taking natural numbers 1, 2,..., n. According to the preview control theory, this function can be minimized by inputting the target reference values for the next N steps:

[0047]

[0048] in, Indicates that the status increases. represents the expansion state at the kth moment, represents the preview ZMP gain, p ref (j) represents the reference ZMP at the jth moment.

[0049] in:

[0050]

[0051] Among them, R represents the penalty coefficient matrix, and T(i-1) represents the i-1th period.

[0052] The matrix P is the solution to the following Riccati equation:

[0053]

[0054] The lateral motion trajectory of the center of mass can be obtained by combining formulas (4), (6), (7), and (8).

[0055] Step S3, using a straight knee optimizer and a stable filter to generate sagittal plane anthropomorphic walking motion, the straight knee optimizer is used to optimize the hip joint trajectory with high fluctuations, and the stable filter is used to generate a hip joint trajectory that meets dynamic constraints.

[0056] The generation of sagittal anthropomorphic walking motion mainly consists of two steps. First, the hip joint trajectories at different heights are optimized using a straight knee optimizer to achieve knee extension features. Second, a stable filter with dynamic constraints is used to generate a stable hip joint trajectory.

[0057] According to the law of human walking, in the single-foot support period, the center of mass height will increase with the walking process, and in the double-foot support period, the supporting legs are mainly exchanged, accompanied by a small change in height. Therefore, in one embodiment, the center point of the single-foot support period and the starting point of the double-foot support period are selected as variables to be optimized, and the hip joint height design problem is expressed as a multi-variable optimization problem of formula (9), and then the fifth-order polynomial is used to connect the points.

[0058]

[0059] Among them, h j (j=1, 2, ..., 2Nstep-1) is the point to be optimized on the hip joint height curve, and Nstep is the total number of steps. hip (i) represents the i-th point in the hip joint height curve. and Indicates the length of the virtual left leg and right leg, that is, the straight-line distance from the hip joint to the ankle joint, Lleg is the actual length of the leg. and h are the upper and lower threshold values of the height. H stand represents the height when standing.

[0060] In the optimization, in order to avoid the situation that the obtained hip joint height is too high and leads to no solution in inverse kinematics, constraints on leg length and height are carried out. In the objective function, soft constraints are introduced to make the height as close as possible to the height in the standing state, and the distance from the center of mass to the end is expected to be as close as possible to the length when the leg is straight. In the inequality constraint, through threshold constraint, while making it reach as high as possible, the distance from the center of mass to the end does not exceed the leg length. As Figure 4 shown, the orange dotted line represents the virtual leg length, and its calculation formula is expressed as:

[0061]

[0062] The stability filter is mainly to make the generated joint trajectory meet the dynamic constraints. For this purpose, the hip joint in the x direction needs to be optimized. First, the reference acceleration value is designed as:

[0063] a d = k p (x d - x) + k d (v d - v) (11)

[0064] where, y hip represents the position of the hip joint on the y-axis, x hip represents the position of the hip joint on the x-axis, x1 angle represents the position of the left ankle joint on the x-axis, z hip represents the position of the hip joint on the z-axis, represents the position of the left ankle joint on the z-axis, x r angle represents the position of the right ankle joint on the x-axis, represents the position of the right ankle joint on the z-axis. k p , k d are control coefficients, x d , v d , a dare the desired position, desired velocity, and desired acceleration of the hip joint, respectively, and x and v represent the actual position and velocity. The optimization goal is to track the desired acceleration and make the generated ZMP track the desired ZMP in the support polygon. Therefore, the objective function consists of several items, namely, the acceleration tracking error, the ZMP tracking error, the quadratic term of acceleration, and the acceleration change. The quadratic term of acceleration ensures moderate acceleration and limits the deviation of position and velocity, consistent with the target tracking acceleration. In addition, the introduction of the acceleration change term can avoid the stabbing motion in the generated gait. In terms of inequality constraints, first, the same geometric virtual leg length constraint as the height optimization is introduced. Secondly, the ZMP is calculated by the center of mass dynamics model, and a stability constraint is added to ensure that the generated ZMP is within the support polygon. Therefore, the optimization problem can be written as:

[0065]

[0066] Among them, W1, W2, W3 are different weight coefficients. and p are the upper and lower thresholds of ZMP, a(i) is the acceleration at the current moment, and a d (i) is the expected acceleration at the current moment, a(i-1) is the acceleration at the previous moment, and p x (i) is the value of ZMP on the X-axis at the current moment, p d (i) is the expected value of ZMP on the X-axis at the current moment.

[0067] Step S4, using the Levenberg-Marquardt algorithm to solve the inverse kinematics to obtain the trajectory of each joint of the robot.

[0068] In order to further verify the effect of the present invention, an experimental verification was carried out. The verification results of the straight knee optimizer are as follows Figure 5 As shown, the blue line represents the height of the hip joint, while the orange and green lines represent the virtual leg length of the exoskeleton. The alternating gray and white backgrounds represent the single support period (SSP) and the double support period (DSP), respectively. The analysis shows that the optimized hip joint height exhibits cyclical changes, reaching a maximum value during the SSP and a minimum value during the DSP. At the same time, the virtual leg length also shows cyclical changes. For example, in the initial cycle, as the hip joint height increases and the gait transitions to the SSP, the right leg plays the role of the swing leg, while the left leg plays the role of the support leg. As the hip joint height continues to rise, the virtual leg length of the support leg approaches the maximum value, while the knee joint simultaneously reaches the minimum angle. Subsequently, as the hip joint height decreases and the gait transitions to the DSP, the virtual leg lengths of the two legs gradually approach each other, and the roles of the support leg and the swing leg are exchanged. These observations are consistent with the characteristics obtained from the analysis of human straight-knee walking motion.

[0069] Figure 6Shows the results of the stability filter, where the dashed line represents the reference value of the trajectory, the gray represents the position reference value, the purple represents the speed reference value, the solid line represents the optimized result, the blue line represents the position, and the orange represents the speed. In the experimental verification, a six-step walking plan was carried out, with each step having a step length of 0.15 m. Starting from Figure 6 It can be seen that the reference value generally shows uniform motion. However, to ensure the overall stability of the robot, the stability filter also has acceleration and deceleration changes while tracking the reference value. For example, during the double-support phase, the speed is relatively fast, and during the single-support phase, the speed is relatively slow. Considering the change of the support polygon during the walking cycle, this acceleration and deceleration law conforms to the human straight-knee walking motion law.

[0070] The simulation results of the anthropomorphic walking gait of the self-balancing lower-limb exoskeleton are as Figure 7 shown. This gait enables the self-balancing lower-limb exoskeleton to have characteristics such as changes in the height of the center of mass, straightening of the knees, heel strike, and toe-off.

[0071] In summary, compared with the prior art, the present invention mainly has the following advantages: decoupling the motion trajectory into the sagittal plane and the coronal plane, generating the sagittal plane motion trajectory on the premise of determining the coronal plane (lateral) motion trajectory, which can ensure the stability of the trajectory; using an optimization method to generate the exoskeleton center of mass coronal plane trajectory, ensuring the stability of the exoskeleton's motion in the coronal plane direction; using a state vector machine and polynomial interpolation method to switch the sole trajectory, which can ensure the smoothness of the trajectory and make the gait have anthropomorphic characteristics; using an optimization method to optimize the exoskeleton center of mass sagittal plane trajectory, ensuring that the knee joint can be straightened as much as possible while satisfying the virtual leg length constraint, making the gait anthropomorphic and satisfying the kinematic constraints; introducing geometric constraints, stability constraints, and acceleration constraints into the coupled trajectory, making the trajectory smooth and stable; using the Levenberg-Marquardt (LM) algorithm to solve the robot inverse kinematics problem, and by adjusting the damping coefficient, the joint singularity problem can be handled.

[0072] In summary, in the prior art, most motion pattern generators decouple motion generation into the sagittal plane and the coronal plane, plan them separately, and then re-couple the trajectories manually. This method usually causes the robot to exceed the final task space when performing knee extension planning due to pursuing a small knee angle. The present invention introduces a virtual leg length geometric constraint to solve this problem. This constraint also kinematically re-couples the motions of the two planes to meet the geometric constraints and ensure the stability of each plane and the requirements of knee extension. In short, the anthropomorphic walking gait generation method for the self-balancing lower-limb exoskeleton provided by the present invention can generate an anthropomorphic walking gait with characteristics such as changes in the height of the center of mass, straightening of the knees, heel strike, and toe-off based on hierarchical optimization, preview control, and a state vector machine.

[0073] The present invention may be a system, a method, and / or a computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions thereon for causing a processor to implement aspects of the present invention.

[0074] The computer-readable storage medium may be a tangible device that can retain and store instructions for use by an instruction execution device. The computer-readable storage medium may be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device such as a punch card or raised structures in a groove having instructions stored thereon, and any suitable combination of the foregoing. The computer-readable storage medium as used herein is not to be construed as a transitory signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through an optical fiber cable), or an electrical signal transmitted through a wire.

[0075] The computer-readable program instructions described herein may be downloaded to each computing / processing device from a computer-readable storage medium or may be downloaded to an external computer or an external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0076] The computer program instructions for performing the operations of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages such as Smalltalk, C++, Python, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider). In some embodiments, by using the state information of the computer-readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the present invention.

[0077] Aspects of the present invention are described herein with reference to the flowchart and / or block diagram of a method, apparatus (system), and computer program product according to an embodiment of the present invention. It should be understood that each block of the flowchart and / or block diagram, and the combination of blocks in the flowchart and / or block diagram, can be implemented by computer-readable program instructions.

[0078] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine such that when these instructions are executed by the processor of the computer or other programmable data processing device, a device is produced that implements the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions may also be stored in a computer-readable storage medium, and these instructions cause a computer, a programmable data processing device, and / or other devices to operate in a specific manner. Thus, the computer-readable medium storing the instructions includes a manufacture that includes instructions for implementing various aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0079] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other devices to produce a computer-implemented process such that the instructions executed on the computer, other programmable data processing apparatus, or other devices implement the functions / acts specified in one or more boxes of the flowchart and / or block diagram.

[0080] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of an instruction, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or acts, or by a combination of dedicated hardware and computer instructions. It is well known to those skilled in the art that implementations by hardware, by software, and by a combination of software and hardware are equivalent.

[0081] The embodiments of the present invention have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of the terms used herein is intended to best explain the principles of the embodiments, the practical application, or the improvement of technologies in the market, or to enable other ordinary skilled persons in the art to understand the embodiments disclosed herein. The scope of the present invention is defined by the appended claims.

Claims

1. A method for generating anthropomorphic walking gait of a self-balancing lower limb exoskeleton, comprising the following steps: According to the walking motion law of human body, the reference zero moment point trajectory, the center of mass sagittal plane trajectory and the sole trajectory are initialized; Based on the initialized reference zero-moment point trajectory, the lateral motion of the self-balancing lower limb rehabilitation exoskeleton is generated using the preview control method. Based on the initialized sagittal trajectory of the center of mass and the sole trajectory, a straight knee optimizer and a stable filter are used to generate a sagittal motion, wherein the straight knee optimizer is used to optimize the hip joint trajectories at different heights, and the stable filter is used to generate a hip joint trajectory that satisfies dynamic constraints; The Levenberg-Marquardt algorithm is used to solve the inverse kinematics and obtain the joint trajectories of the self-balancing lower limb rehabilitation exoskeleton.

2. The method according to claim 1, wherein The straight knee optimizer optimizes the hip joint trajectories at different heights by solving the following objective function: Among them, h j is the point to be optimized on the hip height curve, Nstep is the total number of walking steps, z hip (i) represents the i-th point on the hip height curve, and respectively represent the virtual left leg length and right leg length, L leg is the actual length of the leg, and h are the upper and lower threshold values of the height, H stand represents the height when standing, W1, W2, W3 are different weight coefficients, subject to means subject to.

3. The method according to claim 2, wherein The virtual left leg length and the right leg length are respectively set as: Among them, y hip represents the position of the hip joint on the y-axis, x hip represents the position of the hip joint on the x-axis, x1 angle represents the position of the left ankle joint on the x-axis, z hip represents the position of the hip joint on the z-axis, represents the position of the left ankle joint on the z-axis, x r angle represents the position of the right ankle joint on the x-axis, represents the position of the right ankle joint on the z-axis.

4. The method according to claim 3, wherein The stabilization filter generates a hip joint trajectory that satisfies the dynamic constraints by solving the following objective function: Among them, W1, W2, W3, and W4 are different weight coefficients, and p are the upper and lower thresholds of the ZMP respectively, a(i) is the acceleration at the current moment, a d (i) is the expected acceleration at the current moment, a(i - 1) is the acceleration at the previous moment, p x (i) is the value of the zero moment point on the x-axis at the current moment, p d (i) is the expected value of the zero moment point on the x-axis at the current moment.

5. The method according to claim 4, wherein The desired acceleration is expressed as: a d = k p (x d - x) + k d (v d - v) where k p and k d are control coefficients, x d , v d , a d are the desired position, desired velocity and desired acceleration of the hip joint respectively, and x and v represent the actual position and velocity.

6. The method according to claim 5, wherein During the initialization process, the sole trajectory is determined based on a finite state machine and polynomial interpolation, the state phases include a toe-off phase, a single-foot support phase, and a heel-on phase, and the position of the sole is determined according to the following formula: Among them, and are the position of the ankle joint of the swinging leg in the x-axis direction, the position in the z-axis direction, and the angle on the pitch joint, respectively. L stride is the step length, L forefoot and L heel represent the length of the forefoot and the length of the heel, respectively. θ toe and θ heel represent the pitch angle limits of the toes and the heel, respectively. H ankle represents the height of the ankle joint, H lift represents the maximum height of the foot lift. T c , T to , T hs and T m represent a single gait cycle, the toe-off phase cycle, the heel-strike phase cycle, and the single-support phase cycle, respectively. k represents a natural number, and t represents the moment.

7. The method according to claim 6, characterized in that The lateral motion trajectory of the center of mass is obtained by combining the following equations: where, Δu(i) = u(i) - u(i - 1), ΔΦ(i) = Φ(i) - Φ(i - 1), p y represents the position of the zero moment point on the y-axis, z c is the height of the center of mass in the world coordinate system, g is the acceleration due to gravity, Δt is the sampling time, A represents the dynamic relationship between the internal state variables of the system, B is the input matrix, C is the output matrix, Φ(i + 1) represents the output at the next moment, p y (i) represents the position of the zero moment point in the y-axis direction at the i-th moment, Φ * (i + 1) represents the expanded output at the next moment, Q and R are positive weighting factors, represents the position of the reference zero moment point on the y-axis at the j-th moment, J represents the performance index, and j is the time variable.

8. The method according to claim 1, characterized in that, During initialization, the reference zero moment point trajectory and the sagittal plane center of mass trajectory are determined based on the number of steps, step length, and period.

9. A computer-readable storage medium having a computer program stored thereon, wherein, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer device, comprising a memory and a processor, wherein a computer program capable of running on the processor is stored on the memory, characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images