Tendon relaxation detection method, device and equipment for tendon driving system and medium

Abstract

The invention discloses a tendon relaxation detection method and device for a tendon driving system, equipment and a medium. The method mainly comprises the following main steps that a tendon driving system dynamic model is obtained; and designing a tendon relaxation fault diagnosis observer for judging whether the tendon is relaxed or not. According to the method, through model identification of the tendon driving system, the dynamic characteristics of the tendon driving system can be completely described, and then the tendon relaxation detection observer is designed according to the fault diagnosis principle to capture the tendon relaxation signal. Real-time, robust and low-cost tendon relaxation detection can be realized, the normal work of the tendon driving system is not required to be interfered by shutdown, artificial experience is not required to be relied on, an additional detection unit is not required to be added, the maintenance cost of the tendon driving system is reduced, and the reliability of the tendon driving system is improved.

Description

Technical Field

[0001] The present application relates to the field of robotics, and in particular to a method, device, equipment and medium for detecting tendon relaxation in a tendon drive system. Background Art

[0002] The tendon drive system is a common mechanical transmission device widely used in robotics and biomedical engineering. The movement of the joints is achieved by simulating the pulling function of the muscles through flexible wires (such as steel wire or high-strength fiber). Compared with traditional rigid transmission systems, the tendon drive system uses lightweight materials and is lighter, which reduces energy consumption and improves energy efficiency. In addition, due to the flexible characteristics of the tendon, the drive system has a certain buffering effect when transmitting force, which helps to absorb shock and vibration, reduce mechanical wear, and improve the durability and reliability of the system. The tendon drive system imitates the working principles of biological muscles and tendons, and can achieve movement patterns similar to biological joints. This gives it a unique advantage in the design of bionic robots and prosthetics, and can provide more natural and flexible movements.

[0003] Since the tendon drive system involves the use of flexible materials, its design and maintenance are relatively complex. The tension of the tendon needs to be precisely controlled, and the assembly and adjustment of the system requires a high level of technology. Especially during long-term use, the tendon may become loose or worn, requiring regular inspection and replacement. However, regular inspections are time-consuming and labor-intensive, and it is impossible to detect tendon relaxation in a timely manner or detect the health of the tendon drive system, which reduces the reliability of the tendon drive system. Summary of the invention

[0004] The present application provides a method, device, equipment and medium for detecting tendon relaxation in a tendon drive system, which has the advantage of being able to monitor the relaxation state of the tendon drive system in real time.

[0005] The technical solution of this application is as follows:

[0006] On the one hand, the present application provides a method for detecting tendon relaxation in a tendon drive system, comprising the following steps:

[0007] S1: Obtain the dynamic model of the tendon drive system;

[0008] S2: Establish a tendon relaxation fault diagnosis observer based on the tendon drive system dynamics model to obtain the estimated value of the tendon relaxation detection quantity;

[0009] S3: Compare the estimated value of the tendon relaxation detection amount obtained with a preset range, and determine that the tendon drive system is relaxed when the range is exceeded.

[0010] Further, step S1 includes the following steps:

[0011] S11: Establish the dynamic model of the tendon drive system and organize it into a linear form;

[0012] S12: Tension the tendon drive system and execute a specific trajectory, recording the motor torque, motor position, and joint position during the process;

[0013] S13: Identify the dynamic parameters of the tendon drive system through the dynamic model in step S11 and the data obtained in step S12, and obtain the identified dynamic model of the tendon drive system.

[0014] Furthermore, in step S11, the dynamic model of the tendon drive system is established as:

[0015]

[0016] where τ m Represents the motor driving torque; q m , Respectively represent the motor position, speed and acceleration; δ q It represents the equivalent angle change of tendon deformation; is δ q The derivative with respect to time; I m Indicates the motor rotor inertia; k t represents the equivalent stiffness of the tendon drive system; represents the friction effect of the tendon drive system; τ p represents the pre-tensioning torque bias of the tendon drive system;

[0017] The dynamic model of the tendon-driven system in formula (1) is further organized into the following linear form:

[0018]

[0019] Among them, H t represents the regression matrix, β is the dynamic parameters of the tendon drive system including the motor rotor inertia, the equivalent stiffness of the tendon drive system, the friction parameter and the pre-tensioning torque bias;

[0020] In step S13, the motor torque data, motor position and joint position data obtained at K uniform and continuous moments are stacked and expressed in the following form:

[0021]

[0022] Among them, H t (t k ) indicates t k The regression matrix at time The dynamic parameter identification of the tendon drive system is transformed into the following optimization problem:

[0023]

[0024] Among them I m0 represents the prior value of the motor rotor inertia; α>0 represents the regularization coefficient; when the prior value of the motor rotor inertia is not considered, the parameter β can be estimated by the least squares method, that is,

[0025] Further, step S2 includes the following steps:

[0026] S21: Define the tendon relaxation detection quantity τ s Used to characterize the presence or absence of slack in the tendon drive system:

[0027]

[0028] When the tendon is tensed in the tendon drive system, the tendon relaxation detection value τ s ≈0;

[0029] S21: Establish a tendon relaxation fault diagnosis observer to obtain an estimated value of the tendon relaxation detection amount.

[0030] Furthermore, the tendon relaxation fault diagnosis observer is in the form of:

[0031]

[0032] in is τ s The estimate of k o >0 is the observation gain.

[0033] The tendon relaxation fault diagnosis observer does not need to use acceleration, which avoids the excessive noise introduced by numerical difference and affects the observer performance.

[0034] Furthermore, the tendon relaxation fault diagnosis observer is in the form of:

[0035]

[0036] in Observation gain α 1 ,α 1 ,λ,θ are both greater than zero, represents the generalized momentum observation error, generalized momentum is the estimated value of p, the saturation function sat σ (x) = max{-σ, min{σ, x}}, where σ is the observer feedback error threshold.

[0037] The observer adaptively adjusts the saturation threshold, and the saturation nonlinear link can constrain the abnormal measurement value to cause the observation error to change suddenly, limiting its influence on the observer variable, thereby improving the observer performance. Therefore, the tendon relaxation fault diagnosis observer has a certain robustness to abnormal changes in the motor torque, providing high reliability of tendon relaxation detection.

[0038] Further, in step 3, based on the estimation of the amount of tendon relaxation detected Determine whether there is a tendon relaxation fault in the current tendon drive system:

[0039]

[0040] where ε l , ε u is the preset detection variable. When there is a non-zero value, it is considered that the tendon drive system is relaxed.

[0041] In another aspect, the present application provides a tendon relaxation detection device for a tendon drive system, comprising:

[0042] A tendon drive system dynamics model acquisition unit is used to establish a tendon drive system dynamics model, identify the dynamics parameters of the tendon drive system, and obtain the identified tendon drive system dynamics model;

[0043] A tendon relaxation observation unit is used to establish a tendon relaxation fault diagnosis observer based on a tendon drive system dynamics model to obtain an estimated value of a tendon relaxation detection amount;

[0044] Judgment unit: used to compare the estimated value of the tendon relaxation detection amount obtained with a preset range, and judge that the tendon drive system has relaxed when the limit is exceeded.

[0045] On the other hand, the present application provides a tendon drive system control device, including a processor and a memory, wherein the memory stores computer instructions, and the processor is configured to call and execute the computer instructions to implement the tendon relaxation detection method of the tendon drive system as described above.

[0046] On the other hand, the present application provides a computer-readable medium, wherein the computer-readable medium stores computer instructions, and when the computer instructions are called and executed by a computer, the tendon relaxation detection method of the tendon drive system as described above is implemented.

[0047] In summary, the beneficial effects of this application are:

[0048] 1. The present invention provides a low-cost, high-reliability, easy-to-implement method for detecting tendon relaxation in a tendon drive system. The method converts tendon relaxation detection into a fault diagnosis problem. By establishing a system model and using an observer to estimate the relaxation detection amount, it can be determined whether the tendon drive system has relaxed by judging whether the relaxation detection amount exceeds a set threshold.

[0049] 2. The present invention provides a method for detecting tendon relaxation of a tendon drive system. The method can monitor the relaxation state of the tendon drive system in real time and can realize non-stop detection without relying on regular manual inspection of whether the tendon is relaxed, thereby reducing the maintenance cost of the tendon drive system.

[0050] 3. One of the tendon relaxation fault diagnosis observers provided in the scheme of the present invention does not need to use acceleration, thus avoiding the introduction of excessive noise by numerical difference that affects the performance of the observer; another tendon relaxation fault diagnosis observer has a certain robustness to abnormal changes in motor torque, providing highly reliable tendon relaxation detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a schematic diagram of a tendon drive system;

[0052] Figure 2 This is a schematic diagram of a tendon relaxation detection process in a specific embodiment of the present application;

[0053] Figure 3 is a schematic diagram of a tendon relaxation detection signal at the initial moment in a specific embodiment of the present application;

[0054] Figure 4 It is a schematic diagram of a tendon relaxation detection signal after an endurance test in a specific embodiment of the present application. DETAILED DESCRIPTION

[0055] The specific implementation of the present application is described in detail below with reference to the accompanying drawings.

[0056] Embodiment 1: A method for detecting tendon relaxation in a tendon drive system, comprising the following steps:

[0057] S1: Obtain the dynamic model of the tendon drive system; including:

[0058] S11: Establish the dynamic model of the tendon drive system and organize it into a linear form;

[0059] The dynamic model of the tendon drive system is established as:

[0060]

[0061] where τ m Represents the motor driving torque; q m , Respectively represent the motor position, speed and acceleration; δ q It represents the equivalent angle change of tendon deformation, which can be obtained by combining the encoder readings on the motor side and the joint side with the mechanical parameters of the transmission mechanism; is δ q The derivative with respect to time; I m Indicates the motor rotor inertia; k t represents the equivalent stiffness of the tendon drive system; represents the friction effect of the tendon drive system; τ p represents the pre-tensioning torque bias of the tendon drive system;

[0062] The dynamic model of the tendon-driven system in formula (1) is further organized into the following linear form:

[0063]

[0064] Among them, H t represents the regression matrix, and β is the dynamic parameters of the tendon drive system including the motor rotor inertia, the equivalent stiffness of the tendon drive system, the friction parameter, and the pre-tensioning torque bias.

[0065] S12: Tension the tendon drive system and execute a specific trajectory, recording the motor torque, motor position and joint position during the process, wherein the motor position and joint position are obtained through encoder readings on the motor side and the joint side, respectively.

[0066] S13: Identify the dynamic parameters of the tendon drive system through the dynamic model in step S11 and the data obtained in step S12, and obtain the identified dynamic model of the tendon drive system.

[0067] The motor torque data, motor position and joint position data obtained at K uniform and continuous moments are stacked and expressed as follows:

[0068]

[0069] Among them, H t (t k ) indicates t k The regression matrix at time The dynamic parameter identification of the tendon drive system is transformed into the following optimization problem:

[0070]

[0071] Among them I m0 represents the prior value of the motor rotor inertia, which is provided by the motor manufacturer; α>0 represents the regularization coefficient; when the prior value of the motor rotor inertia is not considered, the parameter β can be estimated by the least square method, that is,

[0072] S2: Establish a tendon relaxation fault diagnosis observer based on the tendon drive system dynamics model to obtain an estimated value of the tendon relaxation detection quantity; including:

[0073] S21: Define the tendon relaxation detection quantity τ s Used to characterize the presence or absence of slack in the tendon drive system:

[0074]

[0075] When the tendon is tensed in the tendon drive system, the tendon relaxation detection value τ s ≈0;

[0076] S21: Establish a tendon relaxation fault diagnosis observer to obtain an estimated value of the tendon relaxation detection amount.

[0077] The tendon relaxation fault diagnosis observer is in the form of:

[0078]

[0079] in is τ s The estimate of k o >0 is the observation gain.

[0080] In some other embodiments of the present application, the tendon relaxation fault diagnosis observer is in the form of:

[0081]

[0082] in Observation gain α 1 ,α 1 ,λ,θ are both greater than zero, represents the generalized momentum observation error, generalized momentum Saturation function sat σ (x) = max{-σ, min{σ, x}}, σ is the observer feedback error threshold. The tendon relaxation fault diagnosis observer has a certain robustness to abnormal changes in motor torque and provides high reliability tendon relaxation detection.

[0083] S3: Compare the estimated value of the tendon relaxation detection amount obtained with a preset range, and determine that the tendon drive system is relaxed when the range is exceeded.

[0084] Estimation of the amount of tendon relaxation measured Determine whether there is a tendon relaxation fault in the current tendon drive system:

[0085]

[0086] where ε l , εu is the preset detection variable. When there is a non-zero value, it is considered that the tendon drive system is loose and needs to be tightened and maintained.

[0087] The following is a detailed description of how to implement tendon relaxation detection using a tendon drive system. In this scenario, the schematic diagram of the tendon drive system is as follows: Figure 1 As shown, the system uses motor M1 as the power source to drive the joint link L1 to move. The transmission mechanism consists of a motor pulley with a radius r, a joint pulley with a radius R, and steel ropes T1 and T2. Motor M1 and link L2 are respectively equipped with encoders to measure angles. Steel ropes L1 and L2 are initially tensioned. First, perform trajectory identification and record the motor torque and dual encoder readings. Use numerical differentiation and phase-free low-pass filtering on the recorded encoder positions to obtain the speed and acceleration on the motor side and joint side. Then solve the optimization problem (4) to obtain the key dynamic parameters of the tendon drive system. Execute in the controller Figure 2 The tendon relaxation detection module shown in the figure first executes module MD1 to obtain the state of the tendon drive system (including motor torque and dual encoder position), and then executes module MD2 based on the identified dynamic parameters and the tendon relaxation detection observer (6), which can calculate the relaxation detection amount estimation in real time. The parameter of this observer is set to α 1 =200,α 2 =10000, λ=10, θ=10, and finally the relaxation judgment module MD3 is executed according to the relaxation detection amount. The result is as follows Figure 3 As shown, the detection threshold ε is set u =0.8,ε l =-0.6, which can ensure the initial tensioning moment No false detection occurs. To verify the effectiveness of the proposed tendon relaxation detection algorithm, a durability test is performed on the tendon drive system to allow the tendon to relax to a certain extent. After a period of durability testing, the estimated results of the relaxation detection amount are as follows: Figure 4 As shown in the figure, between 2 and 5 seconds, Breaking through the set threshold ε l , A relaxation fault signal is triggered, thereby achieving relaxation detection of the tendon system.

[0088] Embodiment 2: A tendon relaxation detection device for a tendon drive system, comprising:

[0089] A tendon drive system dynamics model acquisition unit is used to establish a tendon drive system dynamics model, identify the dynamics parameters of the tendon drive system, and obtain the identified tendon drive system dynamics model;

[0090] A tendon relaxation observation unit is used to establish a tendon relaxation fault diagnosis observer based on a tendon drive system dynamics model to obtain an estimated value of a tendon relaxation detection amount;

[0091] Judgment unit: used to compare the estimated value of the tendon relaxation detection amount obtained with a preset range, and judge that the tendon drive system has relaxed when the limit is exceeded.

[0092] Embodiment 3: A tendon drive system control device comprises a processor and a memory, wherein the memory stores computer instructions, and the processor is configured to call and execute the computer instructions to implement the tendon relaxation detection method of the tendon drive system as described in Embodiment 1.

[0093] Embodiment 4: A computer-readable medium storing computer instructions, wherein when the computer instructions are called and executed by a computer, the method for detecting tendon relaxation in a tendon drive system as described in Embodiment 1 is implemented.

[0094] The above is only a preferred implementation of the present application. It should be pointed out that a person skilled in the art can make several modifications and improvements without departing from the inventive concept of the present application, and these all fall within the scope of protection of the present application.

Claims

1. A method for detecting tendon relaxation in a tendon drive system, characterized in that: The steps include: S1: Obtain the dynamic model of the tendon drive system; S2: Establish a tendon relaxation fault diagnosis observer based on the tendon drive system dynamics model to obtain the estimated value of the tendon relaxation detection quantity; S3: Compare the estimated value of the tendon relaxation detection amount obtained with a preset range, and determine that the tendon drive system is relaxed when the range is exceeded.

2. The method for detecting tendon relaxation of a tendon drive system according to claim 1, characterized in that: Step S1 includes the following steps: S11: Establish the dynamic model of the tendon drive system and organize it into a linear form; S12: Tension the tendon drive system and execute a specific trajectory, recording the motor torque, motor position, and joint position during the process; S13: Identify the dynamic parameters of the tendon drive system through the dynamic model in step S11 and the data obtained in step S12, and obtain the identified dynamic model of the tendon drive system.

3. The method for detecting tendon relaxation of a tendon drive system according to claim 2, characterized in that: In step S11, the dynamic model of the tendon drive system is established as: where τ m Represents the motor driving torque; q m , Respectively represent the motor position, speed and acceleration; δ q It represents the equivalent angle change of tendon deformation; is δ q The derivative with respect to time; I m Indicates the motor rotor inertia; k t represents the equivalent stiffness of the tendon drive system; represents the friction effect of the tendon drive system; τ p represents the pre-tensioning torque bias of the tendon drive system; The dynamic model of the tendon-driven system in formula (1) is further organized into the following linear form: Among them, H t represents the regression matrix, β is the dynamic parameters of the tendon drive system including the motor rotor inertia, the equivalent stiffness of the tendon drive system, the friction parameter and the pre-tensioning torque bias; In step S13, the motor torque data, motor position and joint position data obtained at K uniform and continuous moments are stacked and expressed in the following form: Among them, H t (t k ) indicates t k The regression matrix at time The dynamic parameter identification of the tendon drive system is transformed into the following optimization problem: Among them I m0 represents the prior value of the motor rotor inertia; α>0 represents the regularization coefficient; when the prior value of the motor rotor inertia is not considered, the parameter β can be estimated by the least squares method, that is, 4. The method for detecting tendon relaxation of a tendon drive system according to claim 3, characterized in that: Step S2 includes the following steps: S21: Define the tendon relaxation detection quantity τ s Used to characterize the presence or absence of slack in the tendon drive system: When the tendon is tensed in the tendon drive system, the tendon relaxation detection value τ s ≈0; S21: Establish a tendon relaxation fault diagnosis observer to obtain an estimated value of the tendon relaxation detection amount.

5. The method for detecting tendon relaxation in a tendon drive system according to claim 4, characterized in that: The tendon relaxation fault diagnosis observer is in the form of: in is τ s The estimate of k o >0 is the observation gain.

6. The method for detecting tendon relaxation in a tendon drive system according to claim 4, characterized in that: The tendon relaxation fault diagnosis observer is in the form of: in The observed gains α1, α1, λ, θ are all greater than zero. represents the generalized momentum observation error, generalized momentum is the estimated value of p, is the estimated value of p, the saturation function sat σ (x) = max{-σ, min{σ, x}}, where σ is the observer feedback error threshold.

7. The method for detecting tendon relaxation in a tendon drive system according to claim 5 or 6, characterized in that: In step 3, the estimation of the amount of tendon relaxation is based on the amount of tendon relaxation Determine whether the current tendon drive system has tendon relaxation fault: where ε l , ε u is the preset detection variable. When there is a non-zero value, it is considered that the tendon drive system is relaxed.

8. A tendon relaxation detection device for a tendon drive system, characterized in that: include: A tendon drive system dynamics model acquisition unit is used to establish a tendon drive system dynamics model, identify the dynamics parameters of the tendon drive system, and obtain the identified tendon drive system dynamics model; A tendon relaxation observation unit is used to establish a tendon relaxation fault diagnosis observer based on a tendon drive system dynamics model to obtain an estimated value of a tendon relaxation detection amount; Judgment unit: used to compare the estimated value of the tendon relaxation detection amount obtained with a preset range, and judge that the tendon drive system has relaxed when the limit is exceeded.

9. A tendon drive system control device, characterized in that: It comprises a processor and a memory, wherein the memory stores computer instructions, and the processor is configured to call and execute the computer instructions to implement the tendon relaxation detection method of the tendon drive system as described in any one of claims 1-7.

10. A computer-readable medium, characterized in that The computer-readable medium stores computer instructions, and when the computer instructions are called and executed by a computer, the tendon relaxation detection method of the tendon drive system as described in any one of claims 1-7 is implemented.

Images