Optimization method for passive lower limb assistive exoskeleton joint spring

Abstract

The present application belongs to the field of mechanical design and automation, and particularly relates to a passive lower limb assisting exoskeleton joint spring optimization design method. By introducing the gait of each motion scene in the natural state of the human body and designing it as a corresponding target function as an optimization index, the structure parameters of the passive lower limb assisting exoskeleton joint spring are optimized, and the assisting force effect of the passive lower limb assisting exoskeleton is improved to make it more consistent with human motion. In the optimization of the structure parameters of the passive lower limb assisting exoskeleton joint spring, the steepest descent method is used to obtain the approximate optimal parameters of the exoskeleton joint spring, and the effectiveness of the assisting torque provided by the exoskeleton is improved.

Description

Technical Field

[0001] This invention belongs to the field of mechanical design and automation, and in particular relates to an optimized design method for joint springs of a passive lower limb assistive exoskeleton. Background Technology

[0002] Exoskeletons are wearable devices that provide additional assistance during human movement. They can be categorized by whether they have a power source (active or passive), by power type (pneumatic, hydraulic, motor-driven, or elastically driven), by purpose (assistive, rehabilitation, or enhancement), and by the area they assist (upper limb, lower limb, or full-body exoskeletons).

[0003] In various scenarios such as manufacturing, agriculture, construction, and logistics, workers may encounter situations where they need to climb stairs due to terrain factors. Climbing stairs can cause significant damage to the joints of the lower limbs. Therefore, there is an urgent need for lower limb exoskeleton robots to assist workers in carrying and moving heavy objects in scenarios involving climbing stairs.

[0004] Traditional passive lower limb assistive exoskeletons typically use joint springs to provide assistive torque, achieving an assistive effect in various scenarios. Due to inherent individual differences, gait varies in different scenarios, and the installation parameters of the joint springs directly affect the assistive torque effect of the lower limb assistive exoskeleton. However, current technologies do not address the optimized design of joint spring parameters for passive lower limb assistive exoskeletons. Therefore, optimizing the physical factors of the joint spring installation location has significant application value. Summary of the Invention

[0005] In view of this, the present invention provides an optimized method for the joint spring of a passive lower limb assistive exoskeleton to solve the matching problem between the joint spring auxiliary torque of a traditional passive lower limb assistive exoskeleton and the natural gait of the human body, thereby improving the effectiveness of the auxiliary torque provided by the exoskeleton.

[0006] An optimized method for a passive lower limb assistive exoskeleton joint spring includes the following steps:

[0007] Step 1: Parameter acquisition. The acquired parameters include the angle and torque parameters of the human joints in various motion scenarios under natural conditions, as well as the initial parameters. The natural state refers to the state without wearing the passive lower limb assistive exoskeleton, and the initial parameters refer to the structural parameters of the joint springs after the human wears the passive lower limb assistive exoskeleton.

[0008] Step 2: Establish a dynamic model τ of the joint spring auxiliary torque of the passive lower limb assistive exoskeleton. exo ;

[0009] τ exo =F exo (L1, ..., L) i ,...,u)D[L1,...,L i ,...,u]

[0010] Among them, L i τ represents the optimization parameters; i is the number of optimization parameters; u is the system input of the passive exoskeleton, which is the flexion angle of the human joint; exo The passive lower limb assist exoskeleton provides joint spring-assisted torque; F exo (L1, ..., L) i D[L1, ..., L) represents the auxiliary force of the joint spring in a passive lower limb assistive exoskeleton; i ..., u represents the lever arm length of the joint spring of the passive lower limb assistive exoskeleton. During the algorithm iteration process, u is the angle parameter of the human joint collected in step 1, which is regarded as a calculation constant in the following formula description;

[0011] Step 3: Substitute the collected human joint angle parameters and initial parameters into the dynamic model constructed in Step 2 to obtain the initial auxiliary torque parameters;

[0012] Step 4: Design the objective function for optimizing the joint torque of the human lower limbs:

[0013]

[0014] Where, f(L1, ..., L) i ,...)=τ exo (L1, ..., L) i ,...)-τ hm The objective function is the sum of squared errors between the exoskeleton joint spring torque and the human joint torque; j is the sequence number of the time series corresponding to the collection of human joint parameters in step 1.

[0015] Step 5: Using the initial auxiliary torque parameters obtained in Step 3 as input, the objective function is iteratively calculated and optimized using the steepest descent method to obtain the exoskeleton joint spring torque parameters, so as to improve the matching degree between the joint spring auxiliary torque and the human body's natural gait.

[0016] Furthermore, the angle and torque parameters of human joints in each motion scenario under natural conditions collected in step 1 are the angle and torque parameters of the hip joint, knee joint, or ankle joint.

[0017] Furthermore, the implementation method of step 5 includes the following steps:

[0018] Assuming the iteration reaches the k-th iteration, the objective function is: F = F(L1)(k) , ..., L i (k) , ...), calculate the gradient ΔF of the objective function. i / ΔL i :

[0019] ΔF i / ΔL i =[F(L1, ..., L i +ΔL,...)-F(L1,...,L i ,...)] / ΔL i

[0020] in, ΔF i Let ΔL be the function value during the iteration process. i Here are the parameter values ​​during the iteration process, and c is the learning rate; from point L... i (k) Calculate the next point:

[0021]

[0022] The calculated gradient descent point L i (k+1) Iterate back to ΔF i / ΔL i until the set condition ||ΔF is met. i / ΔL i ||<ε,ε>0,ε is the given precision requirement.

[0023] By adopting the above technical solution, the present invention has the following advantages:

[0024] 1. This invention comprehensively considers practical issues such as the structural parameters of the joint springs of the passive lower limb assistive exoskeleton, introduces the gait of various movement scenarios under the natural state of the human body and designs it as the corresponding objective function as the optimization index, optimizes the structural parameters of the joint springs of the passive lower limb assistive exoskeleton, and makes the passive lower limb assistive exoskeleton more in line with human movement.

[0025] 2. In optimizing the structural parameters of the joint spring of the passive lower limb assistive exoskeleton, the steepest descent method was used to obtain the near-optimal parameters of the exoskeleton joint spring, which improved the effectiveness of the auxiliary torque provided by the exoskeleton. Attached Figure Description

[0026] Figure 1 This is a flowchart of the method for optimizing the joint spring parameters of a passive lower limb assistive exoskeleton in an embodiment;

[0027] Figure 2 This is a schematic diagram illustrating the principle of the passive lower limb assistive exoskeleton joint spring structure in an embodiment. Detailed Implementation

[0028] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0029] like Figure 1 As shown in the figure, this embodiment provides an optimization method for a passive lower limb assistive exoskeleton joint spring, which includes the following steps:

[0030] Step 1: Parameter Acquisition. The acquired parameters include the angle and torque parameters of the human joints in various motion scenarios under natural conditions, as well as initial parameters. Natural conditions refer to the state without wearing the passive lower limb assistive exoskeleton, and initial parameters refer to the structural parameters of the joint springs after the human is wearing the passive lower limb assistive exoskeleton. The collected angle and torque parameters of the human joints in various motion scenarios under natural conditions are the angle and torque parameters of the hip, knee, or ankle joints. This implementation uses the hip joint angle data and hip joint torque information of a human climbing a step under natural conditions.

[0031] Step 2: Construct a dynamic model τ of the joint spring-assisted torque of the passive lower limb assistive exoskeleton. exo The implementation process is as follows:

[0032] Step 2.1, Define τ exo =F exo D l (1);

[0033] Step 2.2, the passive lower limb assistive exoskeleton joint structure used in this embodiment is as follows: Figure 2 As shown, the structure includes a hip joint spring structure rod 1, a joint spring 2, and a thigh structure rod 3. The joint spring structure rod 1 is rotatably connected to the upper end of the thigh structure rod 3. One end of the joint spring 2 is connected to the hip joint spring structure rod 1 and has an initial included angle, while the other end is fixed to the thigh structure rod 3. Based on this passive lower limb assistive exoskeleton joint structure, the following formula can be obtained:

[0034]

[0035] In the formula, k is the spring constant; α is the hip joint flexion angle of the exoskeleton, which is assumed in this embodiment to be consistent with the hip joint flexion angle of the human body; L 3o The spring is the original length. In this embodiment, it is assumed that the initial parameter is the vertical standing state after the human body wears the exoskeleton, and the spring length at this time is the original length; L1 represents the length of the exoskeleton spring structure rod; L2 represents the length of the thigh structure rod.

[0036] Step 3: Substitute the collected human joint angle parameters and initial parameters into formulas (2) and (3), and calculate the initial auxiliary torque parameters according to formula (1);

[0037] Step 4: Design the objective function for optimizing the joint torque of the human lower limbs:

[0038]

[0039] Where, f(L1, ..., Li, ...) = τ exo (L1, ..., L) i ,...)-τ hm The objective function is the sum of squared errors between the exoskeleton joint spring torque and the human joint torque; j is the sequence number of the time series corresponding to the collection of human joint parameters in step 1.

[0040] Step 5: Using the initial auxiliary torque parameters obtained in Step 3 as input, the steepest descent method is used to iteratively calculate and optimize the objective function, obtaining the optimized spring torque parameters to improve the matching degree between the joint spring auxiliary torque and the human body's natural gait. The implementation method is as follows:

[0041] The angle and torque of the human hip joint are presented as a time series, with data from each sampling point input into the optimization model. Assuming the iteration reaches the k-th iteration, the objective function is: F = F(L1) (k) , ..., L i (k) , ...), calculate the gradient ΔF of the objective function. i / ΔL i :

[0042] ΔF i / ΔL i =[F(L1, ..., L i +ΔL,...)-F(L1,...,L i ,...)] / ΔL i

[0043] in, ΔF i Let ΔL be the function value during the iteration process. i Here are the parameter values ​​during the iteration process, and c is the learning rate; from point L... i (k) Calculate the next point:

[0044]

[0045] The calculated gradient descent point L i (k+1) Iterate back to ΔF i / ΔL i until the set condition ||ΔF is met. i / ΔL i||<ε,ε>0,ε is the given precision requirement.

[0046] In summary, the optimization method in this embodiment introduces gait patterns from various movement scenarios under natural human conditions and designs them as corresponding objective functions as optimization indices. Then, using the steepest descent method (Gradient Descent), the structural parameters of the joint springs in the passive lower limb assistive exoskeleton are optimized. This solves the problem of matching the joint spring assist torque of traditional passive lower limb assistive exoskeletons with the natural human gait, making it more consistent with human movement and thus effectively improving the effectiveness of the assist torque provided by the exoskeleton.

[0047] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. An optimized method for a passive lower limb assistive exoskeleton joint spring, characterized in that, Includes the following steps: Step 1: Parameter acquisition. The acquired parameters include the angle and torque parameters of the human joints in various motion scenarios under natural conditions, as well as the initial parameters. The natural state refers to the state without wearing the passive lower limb assistive exoskeleton, and the initial parameters refer to the structural parameters of the joint springs after the human body is wearing the passive lower limb assistive exoskeleton. Step 2: Based on the use of a passive lower limb assistive exoskeleton joint spring structure, construct a dynamic model of the auxiliary torque of the passive lower limb assistive exoskeleton joint spring. ; Among them, L i This represents the optimization parameters; i is the number of optimization parameters; u is the system input for the passive exoskeleton, which is the flexion angle of the human joint. A passive lower limb assist exoskeleton joint spring assist torque; This indicates the auxiliary force of the joint springs in a passive lower limb assistive exoskeleton; The lever arm length of the joint spring of the passive lower limb assistive exoskeleton is represented. In the algorithm iteration process, u is the angle parameter of the human joint collected in step 1, which is regarded as a calculation constant in the following formula description. Step 3: Substitute the collected human joint angle parameters and initial parameters into the dynamic model constructed in Step 2 to obtain the initial auxiliary torque parameters; Step 4: Design the objective function for optimizing the joint torque of the human lower limbs: in, The objective function is the sum of squared errors between the exoskeleton joint spring torque and the human joint torque; j is the sequence number of the time series corresponding to the collection of human joint parameters in step 1. Step 5: Using the initial auxiliary torque parameters obtained in Step 3 as input, the objective function is iteratively calculated and optimized using the steepest descent method to obtain the exoskeleton joint spring torque parameters, so as to improve the matching degree between the joint spring auxiliary torque and the human body's natural gait.

2. The optimized method for a passive lower limb assistive exoskeleton joint spring according to claim 1, characterized in that: The angle and torque parameters of human joints in each motion scenario under natural conditions collected in step 1 are the angle and torque parameters of the hip joint, knee joint, or ankle joint.

3. The optimized method for a passive lower limb assistive exoskeleton joint spring according to claim 2, characterized in that: The implementation method of step 5 includes the following steps: Assuming the iteration reaches the k-th iteration, the objective function is: Calculate the gradient of the objective function. : in, , The function value during the iteration process. Here, c represents the parameter values ​​during the iteration process, and c is the learning rate. From point Calculate the next point: The calculated gradient descent point Re-iterate back until the set conditions are met. , Given the required precision.

Images