Abrasion testing machine for self-monitoring multi-degree-of-freedom movement of artificial knee joint and monitoring method

Abstract

The invention belongs to the technical field of knee joint friction tests, and discloses a self-monitoring artificial knee joint multi-degree-of-freedom motion wear testing machine and a monitoring method.The self-monitoring artificial knee joint multi-degree-of-freedom motion wear testing machine comprises a rack and a testing assembly, and the testing assembly comprises a supporting base, an oil cylinder, a hydraulic rod, a thighbone end assembly and a third motor; a tibia end assembly, a centering platform, a pressure sensor and a sliding platform are sequentially arranged below the thighbone end assembly, a second sliding rail parallel to the supporting base is arranged between the sliding platform and the rack, a monitoring device is arranged at the corner, close to the centering platform, of the rack, and a torsion shaft is sleeved with the pressure sensor. The other end of the torsion shaft is fixedly connected with the first reducer and the first motor; a beam is fixedly connected to the bottom of the sliding platform and connected with a swing part, and the swing part is further connected with a second speed reducer and a second motor. According to the invention, a hydraulic, electric and mechanical combined design mode is adopted, and omnibearing simulation is realized aiming at a real multi-degree-of-freedom motion mode of the knee joint of a human body.

Description

Technical Field

[0001] The invention relates to the technical field of knee joint friction test, and in particular to a wear tester and a monitoring method for self-monitoring multi-degree-of-freedom motion of an artificial knee joint. Background Art

[0002] The artificial knee prosthesis is subject to wear and tear in the human body, resulting in a short service life of the artificial knee joint. Therefore, the wear resistance of the artificial knee prosthesis determines its service life. In order to verify the wear resistance of the artificial knee prosthesis, an in vitro wear evaluation of the artificial knee prosthesis should be carried out. However, the current humanoid knee joint friction and wear testing machines mostly use three-dimensional force sensors or six-dimensional force sensors to monitor multi-degree-of-freedom forces, which have high hardware costs and limited force degrees of freedom. In addition, the installation and calibration of force sensors also make hardware structure design and subsequent maintenance more difficult.

[0003] Chinese invention patent: Publication number is "CN116256262A", and the name is "A knee joint friction test mechanism and its test method", which discloses a knee joint friction test mechanism and its test method, including a knee joint prosthesis, a frame, a six-dimensional force sensor, a slide base plate, a slide fixing plate, a first gait drive device, a transfer shaft, a second gait drive device, a third gait drive device and a controller, and the controller is respectively connected to the six-dimensional force sensor, the first gait drive device, the second gait drive device and the third gait drive device. This technical solution uses the controller to respectively operate the first gait drive device, the second gait drive device and the third gait drive device according to each gait data to complete the femoral flexion gait action, the front and rear displacement gait action and the vertical load gait action, and obtains the mechanical data corresponding to each gait action through the six-dimensional force sensor, thereby achieving the purpose of simulating the complex multi-degree-of-freedom physiological movement of the human knee joint and showing the correct biological movement characteristics of the knee joint. However, this technical solution uses a six-dimensional force sensor, which has a high hardware cost and limited force freedom.

[0004] Chinese invention patent: Publication number is "CN118483090A", and the name is "A four-degree-of-freedom knee prosthesis friction and wear testing machine", which discloses a four-degree-of-freedom knee prosthesis friction and wear testing machine, which includes a frame; a femoral flexion mechanism, which is arranged on the frame and can be vertically lifted and lowered along the frame; the femoral flexion mechanism is used to position the femoral prosthesis and drive the femoral prosthesis to rotate along the vertical plane; the tibial rotation mechanism is located below the tibial prosthesis and is used to carry the tibial prosthesis, and is also used to drive the tibial prosthesis to rotate along the horizontal plane; the front-to-back displacement mechanism is arranged on the frame and supports the tibial rotation mechanism; the front-to-back displacement mechanism is used to drive the tibial rotation mechanism to move forward and backward to drive the tibial prosthesis to move forward and backward relative to the femoral prosthesis; a vertical load application mechanism is arranged on the frame and connected to the bottom of the femoral flexion mechanism; the vertical load application mechanism is used to pull down and release the femoral flexion mechanism to drive the femoral prosthesis to press down and release the tibial prosthesis. This technical solution can achieve four-degree-of-freedom movements, simulate the complex motion state of the human knee joint, and provide effective data for evaluating artificial knee joints. However, this technical solution still uses traditional multi-dimensional force sensors to collect the current force state of the point of application, and cannot perform simulation control that is more in line with the motion law of the human knee joint based on the joint posture information. Summary of the invention

[0005] In order to solve the problem that the conventional multi-dimensional force sensor is used in the above-mentioned prior art to collect the force state of the current application point, and it is not possible to perform simulation control that is more in line with the movement law of the human knee joint according to the joint posture information, and the conventional multi-dimensional force sensor has high hardware cost and limited force freedom. The present invention proposes a wear tester and monitoring method for self-monitoring multi-degree-of-freedom movement of artificial knee joints, so as to solve the problem that the wear tester for artificial knee joints cannot monitor in real time and cannot accurately simulate the complex movement and wear of human knee joints.

[0006] The present invention is realized through the following technical scheme: it comprises a test bench and a test assembly arranged on the test bench, the test assembly comprises a support seat fixedly connected to the test bench, a vertically arranged oil cylinder penetrating the support seat, a hydraulic rod fixedly connected to the oil cylinder, a femoral end assembly fixedly connected to the hydraulic rod, and a third motor perpendicular to the oil cylinder and with an output end fixedly connected to the femoral end assembly, a tibial end assembly, a centering platform, a pressure sensor and a sliding platform are arranged in sequence below the femoral end assembly, the support seat comprises a column and a cross column, one end of the column of the support seat is fixedly connected to the test bench and the other end is fixedly connected to the cross column, the oil cylinder penetrating the cross column of the support seat; One end of the oil cylinder facing the workbench is fixedly connected to the hydraulic rod; a second slide rail parallel to the support seat is arranged between the sliding platform and the workbench, the sliding platform is fixedly connected to the slider of the second slide rail, and the second slide rail is fixedly connected to the workbench; a monitoring device is arranged at a corner of the workbench close to the centering platform, the pressure sensor is sleeved on the torsion shaft, and the other end of the torsion shaft is fixedly connected to the first reducer and the first motor; a crossbeam is fixedly connected to the bottom of the sliding platform, the crossbeam is connected to the swinging member, and the swinging member is also connected to the second reducer and the second motor; the monitoring device, the first motor, the second motor and the third motor are all connected to the computer host.

[0007] As a further preference, the test components are in two groups, which are arranged side by side along the long side of the test stand.

[0008] As a further preference, a crank-connecting rod group is also connected to the torsion shaft, and the crank-connecting rod group is rotationally connected to the torsion shaft on each group of test components.

[0009] As a further preference, a counterweight is arranged between the hydraulic rod and the femoral end assembly; the counterweight is fixedly connected to the hydraulic rod and the femoral end assembly respectively.

[0010] As a further preference, a first slide rail is vertically arranged on the column of the support seat, and a slider of the first slide rail is fixedly connected to the counterweight block.

[0011] The present invention also provides a monitoring method for the wear testing machine for self-monitoring multi-degree-of-freedom motion of an artificial knee joint according to the present invention, comprising the following steps:

[0012] S1. A monitoring device collects and acquires images including components of an artificial knee joint prosthesis;

[0013] S2, taking the pixel content of the knee prosthesis component as the foreground target, using the image target detection algorithm Yolov5 to extract the image of the knee prosthesis component, and setting the background image pixels to 0;

[0014] S3, according to the knee prosthesis component image extracted in step S2, using the pre-trained SAM model as a lightweight image segmentation algorithm to respectively extract images of the tibial end component and the femoral end component;

[0015] S4, according to the images of the tibial component and the femoral component obtained in step S3, using the pre-trained OpenPose model, extract the key points A1, A2, A3, A4, A5, A6 of the femoral component and the key points B1, B2, B3, B4 of the tibial component 3, and obtain the two-dimensional coordinates of the key points of the femoral component and the tibial component;

[0016] S5, according to the two-dimensional coordinates of the key points of the femoral end component and the tibial end component obtained in step S4, the intrinsic parameters K and the extrinsic parameters E of the monitoring device, obtain the posture parameters T=[TA, TB] of the femoral end component and the tibial end component, wherein TA is the homogeneous transformation matrix of the femoral end component with A2 as the reference point, and TB is the homogeneous transformation matrix of the tibial end component with A2 as the reference point;

[0017] S6. The pressure sensor collects the pressure value of the image at the current moment as the force parameter W before posture fine-tuning;

[0018] S7, the posture parameters T of the femoral end component and the tibial end component obtained in step S5 are substituted into the real human knee joint motion model database, and the most similar posture parameters are selected. The similarity calculation formula is:

[0019] argmin||T-Tref||2

[0020] Among them, Tref is the posture parameter of the real human knee joint motion model database, and ||·||2 is the second norm of the matrix.

[0021] S8, according to the most similar posture parameter Tref obtained in step S7 and the posture parameter T of the femoral end component and the tibial end component obtained in step S5, the posture motion parameter is obtained, and the formula is as follows:

[0022] T △ =Tref-T

[0023] The posture motion parameter T △ The force parameter W obtained in the above steps is input into the computer host;

[0024] S9, collecting the pressure value of the pressure sensor after the posture is fine-tuned as the force parameter W' under the posture;

[0025] S10. Repeat steps S1 to S9 until the test is completed.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] 1. The present invention adopts a design method that combines hydraulics, electrics and mechanics to achieve a full range of simulations of the real multi-degree-of-freedom motion of the human knee joint. The hydraulic device can simulate the high-frequency motion of the human knee joint and provide the motion load it bears, thereby achieving a motion method that is closer to the motion characteristics of the human knee joint.

[0028] 2. The present invention adopts an artificial knee prosthesis posture monitoring and control method based on image intelligent algorithm. The posture information of the testing machine is collected by the camera and combined with the collected real motion data of the human knee joint to control and adjust the motor and other equipment, thereby solving the problems of high cost and structural space occupation of traditional multi-dimensional force sensors. It can also provide timely feedback on monitoring results and correct motion tasks to obtain more accurate and comprehensive wear test data.

[0029] 3. The crank-connecting rod mechanism adopted in the present invention can realize multi-degree-of-freedom motion of multiple stations synchronously, thereby improving the test efficiency and reducing the test errors caused by different angles. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the overall structure of the present invention.

[0031] Figure 2 It is a schematic diagram of the structure of a single operating station in the present invention.

[0032] Figure 3 for Figure 2 A partial enlarged schematic diagram of the transmission part.

[0033] Figure 4 for Figure 3 Horizontal side view.

[0034] Figure 5a This is a schematic diagram of an artificial knee joint prosthesis component captured by the monitoring device of the present invention.

[0035] Figure 5b Schematic diagram of respectively extracting the femoral end component and the tibial end component by the lightweight image segmentation algorithm in the present invention.

[0036] Figure 5c It is a schematic diagram of the key point arrangement of the femoral end component and the tibial end component in the present invention.

[0037] Indicated in the figure:

[0038] 1. Cylinder; 2. Femoral end assembly; 3. Tibial end assembly; 4. Centering platform; 5. Pressure sensor; 6. Monitoring device; 7. Sliding platform; 8. First reducer; 9. First motor; 10. Hydraulic rod; 11. First slide rail; 12. Second slide rail; 13. Second reducer; 14. Second motor; 15. Lifting foot; 16. Counterweight; 17. Third motor; 18. Torsion shaft; 19. Crank connecting rod assembly; 20. Crossbeam; 21. Swing assembly. DETAILED DESCRIPTION

[0039] Advantages and features of the present invention will be illustrated and explained by the following non-limiting description of preferred embodiments thereof, which are given by way of example only with reference to the accompanying drawings.

[0040] like Figures 1 to 4As shown, the present invention provides a wear tester for self-monitoring multi-degree-of-freedom motion of artificial knee joint, including a bench, on which a test assembly is arranged, and the test assembly includes a support seat fixedly connected to the bench. The support seat includes a column and a cross column, one end of the column of the support seat is fixedly connected to the bench and the other end is fixedly connected to the cross column, and the support seat preferably has four columns, which are respectively arranged at the four corners of the cross column. The test assembly also includes a cylinder 1, a femoral end assembly 2, a tibial end assembly 3, a centering platform 4, a pressure sensor 5, a sliding platform 7 and a third motor 17. The cross column of the support seat is penetrated by a cylinder 1, and the cylinder 1 is arranged vertically. The cylinder 1 is fixedly connected to a hydraulic rod 10 at one end of the bench, and the other end of the hydraulic rod 10 is fixedly connected to the femoral end assembly 2. The cylinder 1 drives the hydraulic rod 10 to move up and down, which is used to pressurize the femoral end assembly 2, and is used to simulate different human body weights to achieve pressurization actions of different frequencies. A tibial end component 3 is arranged below the femoral end component 2, a centering platform 4 is fixedly connected below the tibial end component 3, a pressure sensor 5 is fixedly connected below the centering platform 4, and a sliding platform 7 is arranged below the pressure sensor 5. A second slide rail 12 is arranged between the sliding platform 7 and the stand, the sliding platform 7 is fixedly connected to the slider of the second slide rail 12, and the second slide rail 12 is fixedly connected to the stand. The second slide rail 12 is arranged parallel to the support seat. The centering platform 4 is used for the upper and lower centering of the tibial end component 3, and the pressure sensor 5 is used for sensing the pressure value exerted on the tibial end component 3. The sliding platform 7 and the second slide rail 12 constitute a horizontal motion platform for the tibial end component 3. The test components are preferably in two groups, which are arranged side by side along the long side direction of the stand. A monitoring device 6 is provided at a corner of the test stand near the centering platform 4. The monitoring device 6 is used to collect image data of the test machine and send it to the computer host. The computer host uses an image intelligent algorithm to identify the posture of the artificial knee joint prosthesis, which is used to monitor the operating status of the test machine in real time, and send instructions to the motor control system to correct the motion task in real time to obtain comprehensive and accurate wear test data and achieve accurate wear results of the artificial knee joint prosthesis. The pressure sensor 5 is sleeved on the torsion shaft 18, and the other end of the torsion shaft 18 is fixedly connected to the first reducer 8 and the first motor 9. The first reducer 8 is arranged above the first motor 9 to match the rotation speed and transmit torque, and serves as a torsional power device for the tibial end assembly 3. In order to facilitate the adjustment of the horizontality of the entire stand, lifting feet 15 are also provided at the four corners of the bottom of the stand, and the lifting feet 15 are fixedly connected to the stand by bolts.

[0041] When there are multiple groups of test components, a crank-connecting rod group 19 is also connected to the torsion shaft 18, and the crank-connecting rod group 19 is rotationally connected to the torsion shaft 18 on each group of test components. The advantage of this design is that the first reducer 8 and the first motor 9 can be set on only one group of test components, and the crank-connecting rod group 19 can be used to drive multiple groups of tibial end components 3 to twist.

[0042] The bottom of the sliding platform 7 is fixedly connected with a crossbeam 20, and the crossbeam 20 is connected to a swinging member 21, and the swinging member 21 is also connected to the second reducer 13 and the second motor 14. The second motor 14 and the second reducer 13 drive the swinging member 21 to swing, thereby driving the crossbeam 20 to move, and finally driving the sliding platform 7 to move horizontally under the restriction of the second slide rail 12.

[0043] like Figure 2 As shown, in order to simulate more smoothly and reduce the load of the oil cylinder 1, a counterweight 16 can be set between the hydraulic rod 10 and the femoral end assembly 2. The counterweight 16 is fixedly connected to the hydraulic rod 10 and the femoral end assembly 2 respectively. At the same time, a first slide rail 11 is vertically arranged on the column of the support seat, and the slider of the first slide rail 11 is fixedly connected to the counterweight 16. The oil cylinder 1 drives the counterweight 16 connected to the hydraulic rod 10 to move up and down, which is used to pressurize the femoral end assembly 2, simulate different human body weights, and realize pressurization actions of different frequencies. The output end of the third motor 17 is fixedly connected to the femoral end assembly 2, and is used to drive the femoral end assembly 2 to rotate, forming a swing mechanism, which is used to simulate knee flexion at different angles of the human body. The third motor 17 is arranged perpendicular to the oil cylinder 1.

[0044] The first motor 9, the second motor 14 and the third motor 17 are all connected to the computer host to receive control signals from the computer host. The femoral end component 2 and the tibial end component 3 are driven by the first motor 9, the second motor 14 and the third motor 17 to achieve multi-degree-of-freedom motion simulating the human knee joint.

[0045] The present invention also provides a monitoring method for the wear testing machine for self-monitoring multi-degree-of-freedom motion of an artificial knee joint according to the present invention, comprising the following steps:

[0046] Step 1: The monitoring device 6 collects and acquires images containing artificial knee joint prosthesis components.

[0047] like Figure 5a , an image of an artificial knee joint prosthesis component acquired by using the monitoring device 6 is shown.

[0048] Step 2: Take the pixel content of the knee prosthesis component as the foreground target, use the image target detection algorithm Yolov5 to extract the image of the knee prosthesis component, and set the background image pixels to 0.

[0049] Step 3: Based on the knee prosthesis component images extracted in step 2, the images of the tibial end component 3 and the femoral end component 2 are respectively extracted using the pre-trained cropped SegmentAnything Model, i.e., the SAM model, as a lightweight image segmentation algorithm.

[0050] like Figure 5b Shown are images of the femoral end component 2 and the tibial end component 3 extracted through the SAM model.

[0051] Step 4. Based on the images of the tibial component 3 and the femoral component 2 obtained in step 3, use the pre-trained OpenPose model to extract the key points A1, A2, A3, A4, A5, A6 of the femoral component 2 and the key points B1, B2, B3, B4 of the tibial component 3, and obtain the two-dimensional coordinates of the key points of the femoral component 2 and the tibial component 3.

[0052] like Figure 5c As shown, there are six key points on the femoral component 2, namely A1: the center away from one end of the tibial component 3; A2: the center close to one end of the tibial component 3; A3, A4: the centers of the left and right side walls; A5, A6: two protrusions on the left and right sides of one end of the tibial component 3. There are four key points on the tibial component 3, namely B1, B2: the left and right ends; B3: the center away from one end of the femoral component 2; B4: the center close to one end of the femoral component 2.

[0053] Step 5. According to the two-dimensional coordinates of the key points of the femoral end component 2 and the tibial end component 3 obtained in step 4, the intrinsic parameters K and extrinsic parameters E of the monitoring device 6, obtain the posture parameters T = [TA, TB] of the femoral end component 2 and the tibial end component 3, where TA is the homogeneous transformation matrix of the femoral end component 2 with A2 as the reference point, and TB is the homogeneous transformation matrix of the tibial end component 3 with A2 as the reference point.

[0054] Step 51, measure and obtain the internal parameter K and external parameter E of the monitoring device 6, the formula is as follows:

[0055]

[0056] Among them, f x and f y is the focal length in the x and y directions, c x and c y are the principal point coordinates, R is the 3x3 rotation matrix, and t is the 3x1 translation vector.

[0057] Step 52, converting the two-dimensional coordinates of the key points of the femoral end component 2 obtained in step 4 into three-dimensional coordinates;

[0058] Step 521, convert the two-dimensional coordinates into homogeneous coordinates, the formula is as follows:

[0059]

[0060] Wherein, i = 1, 2, 3, 4, 5, 6; and They respectively represent the homogeneous coordinates of a key point before and after the change.

[0061] Step 522: According to the intrinsic parameter K of the monitoring device 6 measured in step 51 and the homogeneous coordinates obtained in step 521, the normalized image plane coordinates are obtained, and the formula is as follows:

[0062]

[0063] Among them, u i norm and u i ' norm Respectively represent the normalized image plane coordinates of a key point before and after the change; K -1 is the inverse matrix of the intrinsic parameter K.

[0064]

[0065] Step 523: Obtain the world coordinates according to the normalized image plane obtained in step 522 and the external parameters E of the monitoring device 6 measured in step 51. The formula is as follows:

[0066]

[0067] in is the transpose of the inverse matrix of the external parameter E; x i and x i 'Represents the world coordinates of a key point before and after the change.

[0068] Step 524: Obtain the camera coordinates according to the world coordinates obtained in step 523 and the external parameters E of the monitoring device 6 measured in step 51. The formula is as follows:

[0069]

[0070] in and Respectively represent the camera coordinates of a key point before and after the change.

[0071] Step 525: According to the camera coordinates obtained in step 524 and the inverse matrix of the external parameters E, the three-dimensional coordinates of the object coordinate system are obtained. The formula is as follows:

[0072]

[0073] Where P i and Pi 'Represents the three-dimensional coordinates of the object coordinate system of a key point before and after the change.

[0074] Step 53: Obtain a rotation matrix relative to the reference point A2 according to the three-dimensional coordinates of the object coordinate system obtained in step 525.

[0075] Step 531: According to the three-dimensional coordinates of the object coordinate system obtained in step 225, the coordinates of each key point relative to the reference point A2 are obtained, and the formula is as follows:

[0076]

[0077] Among them, P i rel and P i ' rel They respectively represent the coordinates of a key point before and after the change relative to the reference point A2; P2 and P2′ are the three-dimensional coordinates of the key point A2 obtained in step 525 in the object coordinate system.

[0078] Step 532: P obtained in step 531 i rel and P i ’rel , using singular value decomposition SVD, we get the rotation matrix relative to the reference point A2, the formula is as follows:

[0079]

[0080] Step 54: According to the rotation matrix R relative to the reference point A2 rel The translation vector relative to the reference point A2 is obtained by combining the three-dimensional coordinates of the key point A2 obtained in step 525 in the object coordinate system, and the formula is as follows:

[0081] t rel =p′2-R rel p2

[0082] Step 55, combining the rotation and translation to obtain the homogeneous transformation matrix TA of the femoral end component 2 with A2 as the reference point.

[0083] Step 551: construct a homogeneous transformation matrix relative to reference point A2:

[0084]

[0085] Step 552: Apply the homogeneous transformation matrix relative to the reference point A2 to obtain the TA of the homogeneous transformation matrix of the femoral end component 2 with A2 as the reference point. The formula is as follows:

[0086]

[0087] p′ rel =T rel p hom

[0088] TA=T rel

[0089] Where P is a point on the object; P hom is the homogeneous coordinate of point P; p′ rel are the coordinates of the point after transformation relative to the reference point A2.

[0090] Step 56, replace the key points of the femoral end component 2 in step 52 with the key points of the tibial end component 3 and the key point A2 in the femoral end component 2, repeat steps 52 to 55, and obtain the homogeneous transformation matrix TB of the tibial end component 3 with A2 as the reference point.

[0091] Step 6: The pressure sensor 5 collects the pressure value of the image at the current moment as the force parameter W before posture fine-tuning.

[0092] Step 7: Substitute the posture parameters T of the femoral end component 2 and the tibial end component 3 obtained in step 5 into the real human knee joint motion model database, and select the most similar posture parameters. The similarity calculation formula is argmin||T-Tref||2.

[0093] Among them, Tref is the posture parameter of the real human knee joint motion model database, and ||·||2 is the second norm of the matrix.

[0094] Step 8: According to the most similar posture parameter Tref obtained in step 7 and the posture parameter T of the femoral end component 2 and the tibial end component 3 obtained in step 5, the posture motion parameter is obtained, and the formula is as follows:

[0095] T △ =Tref-T

[0096] The posture motion parameter T △ The force parameter W obtained in step 6 is input into the computer host to realize automatic fine-tuning of the machine.

[0097] Step 9: Collect the pressure value of the pressure sensor 5 after the posture is fine-tuned as the force parameter W' in this posture.

[0098] Step 10: Repeat steps 1 to 9 until the test is completed, that is, the number of tests reaches a set value or the force parameter W' exceeds a set threshold.

Claims

1. A wear tester for self-monitoring multi-degree-of-freedom motion of an artificial knee joint, comprising a stand and a test assembly arranged on the stand, characterized in that: The test assembly comprises a support seat fixedly connected to the bench, a cylinder (1) arranged vertically and passing through the support seat, a hydraulic rod (10) fixedly connected to the cylinder (1), a femoral end assembly (2) fixedly connected to the hydraulic rod (10), and a third motor (17) perpendicular to the cylinder (1) and having an output end fixedly connected to the femoral end assembly (2); a tibial end assembly (3), a centering platform (4), a pressure sensor (5) and a sliding platform (7) are arranged in sequence below the femoral end assembly (2); the support seat comprises a column and a cross column, one end of the column of the support seat is fixedly connected to the bench and the other end is fixedly connected to the cross column, and the cylinder (1) passes through the cross column of the support seat; the end of the cylinder (1) facing the bench is fixedly connected to the hydraulic rod (10); the sliding platform (7) is connected to the bench. A second slide rail (12) is provided parallel to the support seat, the sliding platform (7) is fixedly connected to the slider of the second slide rail (12), and the second slide rail (12) is fixedly connected to the stand; a monitoring device (6) is provided on a corner of the stand close to the centering platform (4), the pressure sensor (5) is sleeved on the torsion shaft (18), and the other end of the torsion shaft (18) is fixedly connected to the first reducer (8) and the first motor (9); a crossbeam (20) is fixedly connected to the bottom of the sliding platform (7), the crossbeam (20) is connected to the swing member (21), and the swing member (21) is also connected to the second reducer (13) and the second motor (14); the monitoring device (6), the first motor (9), the second motor (14) and the third motor (17) are all connected to a computer host.

2. The wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 1, characterized in that: The test components are divided into two groups and arranged side by side along the long side of the test stand.

3. The wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 2, characterized in that: A crank-connecting rod group (19) is also connected to the torsion shaft (18), and the crank-connecting rod group (19) is rotationally connected to the torsion shaft (18) on each test assembly group.

4. The wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 1, characterized in that: A counterweight block (16) is arranged between the hydraulic rod (10) and the femoral end component (2); the counterweight block (16) is fixedly connected to the hydraulic rod (10) and the femoral end component (2) respectively.

5. The wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 4, characterized in that: A first slide rail (11) is vertically arranged on the column of the support seat, and a slider of the first slide rail (11) is fixedly connected to a counterweight block (16).

6. The wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 1, characterized in that: The four corners of the bottom of the platform are also provided with lifting feet (15), and the lifting feet (15) are fixedly connected to the platform by bolts.

7. The wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 1, characterized in that: The support seat has four upright posts, which are respectively arranged at the four corners of the horizontal post.

8. A monitoring method applicable to the wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joints according to any one of claims 1 to 7, characterized in that: The following steps are included: S1, a monitoring device (6) collects and acquires images including components of an artificial knee joint prosthesis; S2, taking the pixel content of the knee prosthesis component as the foreground target, using the image target detection algorithm Yolov5 to extract the image of the knee prosthesis component, and setting the background image pixels to 0; S3, according to the knee prosthesis component images extracted in step S2, using the pre-trained SAM model as a lightweight image segmentation algorithm to respectively extract images of the tibial end component (3) and the femoral end component (2); S4, according to the images of the tibial component (3) and the femoral component (2) obtained in step S3, using the pre-trained OpenPose model, extract the key points A1, A2, A3, A4, A5, A6 of the femoral component (2) and the key points B1, B2, B3, B4 of the tibial component 3, and obtain the two-dimensional coordinates of the key points of the femoral component (2) and the tibial component (3); S5, according to the two-dimensional coordinates of the key points of the femoral end component (2) and the tibial end component (3) obtained in step S4, the internal parameters K and the external parameters E of the monitoring device (6), obtain the posture parameters T = [TA, TB] of the femoral end component (2) and the tibial end component (3), wherein TA is the homogeneous transformation matrix of the femoral end component (2) with A2 as the reference point, and TB is the homogeneous transformation matrix of the tibial end component (3) with A2 as the reference point; S6, the pressure sensor (5) collects the pressure value of the image at the current moment as the force parameter W before the posture fine-tuning; S7, the posture parameters T of the femoral end component (2) and the tibial end component (3) obtained in step S5 are substituted into the real human knee joint motion model database, and the most similar posture parameters are selected. The similarity calculation formula is: argmin||T-Tref||2 Among them, Tref is the posture parameter of the real human knee joint motion model database, and ||·||2 is the second norm of the matrix. S8, according to the most similar posture parameter Tref obtained in step S7 and the posture parameter T of the femoral end component (2) and the tibial end component (3) obtained in step S5, the posture motion parameter is obtained, and the formula is as follows: T △ =Tref-T The posture motion parameter T △ The force parameter W obtained in step (6) is input into the computer host; S9, collecting the pressure value of the pressure sensor (5) after the posture is fine-tuned as the force parameter W' under the posture; S10. Repeat steps S1 to S9 until the test is completed.

9. The monitoring method of the wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 8, characterized in that: The specific steps of step S5 are as follows: S51. Measure and obtain the internal parameter K and external parameter E of the monitoring device (6), the formula is as follows: Among them, f x and f y is the focal length in the x and y directions, c x and c y are the principal point coordinates, R is the 3x3 rotation matrix, and t is the 3x1 translation vector; S52, converting the two-dimensional coordinates of the key points of the femoral end component (2) into three-dimensional coordinates; S521. Convert the two-dimensional coordinates into homogeneous coordinates. The formula is as follows: Wherein, i = 1, 2, 3, 4, 5, 6; and Respectively represent the homogeneous coordinates of a key point before and after the change; S522, according to the intrinsic parameter K and the homogeneous coordinates obtained in step S521, obtain the normalized image plane coordinates, the formula is as follows: Among them, u i norm and u i ' norm Respectively represent the normalized image plane coordinates of a key point before and after the change; K -1 is the inverse matrix of the internal parameter K; S523. Obtain the world coordinates according to the normalized image plane and the external parameter E. The formula is as follows: in is the transpose of the inverse matrix of the external parameter E; x i and x' i Respectively represent the world coordinates of a key point before and after the change; S524. Obtain the camera coordinates according to the world coordinates and the external parameters E. The formula is as follows: in and Respectively represent the camera coordinates of a key point before and after the change; S525. According to the camera coordinates and the inverse matrix of the external parameter E, the three-dimensional coordinates of the object coordinate system are obtained. The formula is as follows: Where P i and P i 'Represents the three-dimensional coordinates of the object coordinate system of a key point before and after the change; S53, obtaining a rotation matrix relative to the reference point A2 according to the three-dimensional coordinates of the object coordinate system; S531. According to the three-dimensional coordinates of the object coordinate system, the coordinates of each key point relative to the reference point A2 are obtained. The formula is as follows: Among them, P i rel and P i ' rel Respectively represent the coordinates of a key point before and after the change relative to the reference point A2; P2 and P2' are the three-dimensional coordinates of the key point A2 obtained in step 525 in the object coordinate system; S532, according to P i rel and P i ' rel , using singular value decomposition SVD, we get the rotation matrix relative to the reference point A2, the formula is as follows: S54, according to the rotation matrix R rel The translation vector relative to the reference point A2 is obtained by combining the three-dimensional coordinates of the key point A2 obtained in step S525 in the object coordinate system, and the formula is as follows: t rel =p′2-R rel p2 S55, combining the rotation and translation to obtain TA of the homogeneous transformation matrix of the femoral end component (2) with A2 as the reference point; S551. Construct a homogeneous transformation matrix relative to the reference point A2: S552, applying the homogeneous transformation matrix relative to the reference point A2, obtaining the TA of the homogeneous transformation matrix of the femoral end component (2) with A2 as the reference point, the formula is as follows: p′ rel =T rel p hom TA=T rel Where P is a point on the object; P hom is the homogeneous coordinate of point P; p′ rel are the coordinates of the point transformed relative to the reference point A2; S56. Replace the key points of the femoral end component (2) in step S52 with the key points of the tibial end component (3) and the key point A2 in the femoral end component (2), repeat steps S52 to S55, and obtain the homogeneous transformation matrix TB of the tibial end component (3) with A2 as the reference point.

10. The monitoring method of the wear testing machine for self-monitoring multi-degree-of-freedom motion of artificial knee joint according to claim 9, characterized in that: The key points of the femoral end component (2) in step S4 are A1: the center away from one end of the tibial end component (3); A2: the center close to one end of the tibial end component (3); A3, A4: the centers of the left and right side walls; A5, A6: two protrusions on the left and right sides of one end of the tibial end component (3); the key points of the tibial end component (3) are B1, B2: the left and right ends; B3: far from the center of one end of the femoral end component (2); B4: close to the center of one end of the femoral end component (2).

Images