An adaptive split-belt treadmill system for adjusting belt speeds

Abstract

The application discloses a self-adaptive adjusting split-belt dual-track treadmill belt speed system, which comprises an information acquisition module, an information analysis module, a controller module, a speed control module and a visual feedback module; the information acquisition module further comprises a motion signal acquisition unit and a pressure signal acquisition unit. The application calculates the average mapping error and the error average change between the real-time symmetry coefficient and the target symmetry coefficient in the training process of the patient in the information analysis module through the kinematic information and the dynamic information collected by the information acquisition module in real time, analyzes the training difficulty change required by the patient at present according to the fuzzy logic controller of the controller module, controls the running speed of the two belts in the speed control module respectively and individually based on the training difficulty change, and provides the patient with the visual feedback with implicit distortion as the training target in the visual feedback module, so as to improve the symmetry and improve the walking ability of the patient.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical rehabilitation, and particularly relates to a self-adaptive speed adjustment system for split-belt treadmill tracks. BACKGROUND

[0002] Due to the impairment of the neural pathway controlling gait, hemiplegic patients have long-term loss of lower limb function, resulting in pathological phenomena such as atrophy of multiple muscles on the hemiplegic side, muscle strength reduction, muscle spasticity, inability of the knee joint to complete flexion and extension, foot drop gait, and loss of body balance stability. Therefore, lower limb motor function rehabilitation has become a hot spot in the clinical rehabilitation of stroke patients. The symmetry of human body movement is an important feature of normal human gait, and the gait pattern of healthy individuals in terms of time, distance and vertical force is quite symmetrical, with little deviation from complete symmetry. However, in stroke patients, there is obvious asymmetry between the hemiplegic side and the non-hemiplegic side, such as shorter support time, longer step length, and smaller ground reaction force on the hemiplegic side relative to the non-hemiplegic side. This asymmetry increases the walking cost of hemiplegic patients and increases the risk of falling during walking. Therefore, improving asymmetry will help reduce the energy consumption of patients during daily walking and enhance the walking ability of patients.

[0003] Split-belt treadmills can independently adjust the speed of the two tracks, and can obtain the kinetic data of the lower limbs of the subject during walking. Reisman et al. found that when walking at different speeds (e.g., on a split-belt treadmill), returning to walking on the ground can reduce step length asymmetry. Helm et al. compared the training effects of patients on a split-belt treadmill at constant speed ratio and variable speed ratio, and found that the training effect of patients at variable speed ratio would be greater. Repeated training on a split-belt treadmill can help patients develop motor learning and improve post-stroke gait. Kim et al. pointed out that providing implicit distorted visual feedback in split-belt treadmill training will help retain training effects to a greater extent. In summary, reasonable regulation of the speed of the two tracks in split-belt treadmill training and the combination of visual feedback will help improve the improvement effect of training and optimize the walking ability of patients.

[0004] Existing research generally uses the independent speed of a split-belt treadmill to adaptively train patients, i.e., to explore the effect of variable speed and split-belt treadmill speed on patient gait. Few studies adjust the speed of the two tracks of a split-belt treadmill according to the gait walking characteristics of patients. Previous studies have shown that gait symmetry is closely related to the speed of the two tracks and their difference, so this patent adjusts the speed of the two tracks of a split-belt treadmill according to the symmetry of the patient's walking process, thereby improving the symmetry of the patient's gait. Summary of the Invention

[0005] The main purpose of this application is to overcome the shortcomings and deficiencies of the prior art and provide an adaptive adjustment system for the speed of a dual-track treadmill belt. Based on the change in training difficulty, the running speed of the two belts is controlled separately, and the visual feedback module provides the patient with implicitly distorted visual feedback as a training target, thereby improving symmetry and enhancing the patient's walking ability.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] This application provides an adaptive adjustment system for the belt speed of a split-belt dual-track treadmill, comprising: an information acquisition module, an information analysis module, a controller module, a speed control module, and a visual feedback module;

[0008] The information acquisition module is used to acquire kinematic and dynamic information of the patient's lower limb joints in real time during the patient's walking training.

[0009] The information analysis module is used to calculate the average mapping error and the average change of error between the real-time symmetry coefficient of the patient's lower limb and the target symmetry coefficient based on the kinematic and dynamic information acquired in real time by the information acquisition module.

[0010] The controller module is a fuzzy logic controller, which is used to deduce the appropriate training difficulty change for the patient based on the average mapping error and the average change of error using fuzzy rules. After each training period, the fuzzy logic controller obtains the patient's parameter changes and calculates the required training difficulty change for the patient in the next period according to fuzzy rules.

[0011] The speed control module is used to adjust the input frequency of the treadmill motor according to the change in training difficulty, and to adjust the running speed of the two treadmill belts separately.

[0012] The visual feedback module is used to provide patients with real-time gait data and training targets that contain implicit distortions, allowing patients to change their own gait characteristics.

[0013] As a preferred technical solution, the information acquisition module further includes a motion signal acquisition unit and a pressure signal acquisition unit;

[0014] The motion signal acquisition unit is used to collect kinematic information of each joint of the patient's lower limbs during movement in real time;

[0015] The pressure signal acquisition unit is used to collect the three-dimensional reaction force between the patient and the track during walking in real time.

[0016] As a preferred technical solution, the information analysis module is specifically used to input the real-time acquired kinematic and dynamic information into the symmetry assessment model during the patient's walking process, and analyze the changes in the error parameters between the patient's real-time symmetry coefficient and the target symmetry coefficient, including the following steps:

[0017] After removing baseline values ​​and low-pass filtering from the real-time acquired dynamic information, gait events are detected and gait cycles are divided using a threshold method.

[0018] The patient's symmetry coefficient within the current gait cycle is calculated using real-time acquired kinematic information.

[0019] By comparing the real-time symmetry coefficient with the target symmetry coefficient, the error between the two in the current gait cycle is obtained;

[0020] Compare the error of the current gait cycle with the error of the previous gait cycle to obtain the current error change.

[0021] The error of the current gait cycle is adjusted by nonlinear mapping to obtain the mapped error;

[0022] The average value of the mapping error and the average change in error of the patient in multiple gait cycles within a training phase are calculated to obtain the average mapping error and the average change in error of the patient within that training phase.

[0023] As a preferred technical solution, the symmetry coefficient is the step-size symmetry coefficient, and the calculation formula is:

[0024]

[0025] Among them, SLA real-time It is the stride length symmetry coefficient of the patient's current gait cycle, SL paretic and SL non-paretic These represent the stride length on the hemiplegic side and the stride length on the non-hemiplegic side during the patient's current gait cycle.

[0026] As a preferred technical solution, the error of the current gait cycle is adjusted by nonlinear mapping to obtain the mapped error. The specific calculation formula is as follows:

[0027]

[0028] Among them, E r is the mapping error of the patient's current gait cycle, E is the error between the real-time step length symmetry coefficient and the target step length symmetry coefficient of the patient's current gait cycle, m is the adjustment parameter for the steepness of the curve in the nonlinear transformation, and b is the adjustment parameter for the mapping range of E in the nonlinear transformation.

[0029] As a preferred technical solution, the controller module is a fuzzy logic controller, and the control process includes fuzzification, fuzzy inference, and defuzzification;

[0030] The fuzzification refers to converting the average mapping error and the average change in error into linguistic variables, and obtaining the membership degree of the average mapping error and the average change in error to the fuzzy set according to the membership function.

[0031] The fuzzy reasoning refers to deriving fuzzy conclusions based on the input membership degree and preset fuzzy rules;

[0032] Defuzzification refers to using defuzzification methods to map fuzzy conclusions to specific non-fuzzy outputs, i.e., the change in training difficulty currently required by the patient.

[0033] As a preferred technical solution, the membership function is a trigonometric membership function;

[0034] The defuzzification method is the region center method, which uses the center of the area enclosed by the membership function curve of the output fuzzy set and the horizontal axis as the output of the fuzzy logic controller, and then converts the output of the fuzzy logic controller into the change in training difficulty required by the patient.

[0035] As a preferred technical solution, the speed control module is specifically used to adjust the motor input frequency of the two belts of the split-belt dual-track treadmill according to the change in training difficulty required by the patient, and to limit the ratio of the motor input frequencies of the two belts to 1:3, thereby independently adjusting the running speed of the two belts of the treadmill. The calculation formula is as follows:

[0036]

[0037] f paretic =f paretic '-rΔD

[0038] f non-paretic =f non-paretic '+rΔD

[0039] v paretic =sf paretic

[0040] v non-paretic =sf non-paretic

[0041] Among them, f paretic with f non-paretic These are the motor input frequencies of the tracks located on the hemiplegic and non-hemiplegic sides during the patient's next training phase, f paretic 'and f non-paretic 'These are the motor input frequencies of the tracks on the hemiplegic and non-hemiplegic sides during the patient's previous training phase, vparetic With v non-paretic These are the running speeds of the treadmill on the hemiplegic and non-hemiplegic sides of the patient's current gait cycle, respectively; r is the transformation coefficient between the change in training difficulty ΔD required by the patient and the input frequency; and s is the transformation coefficient between the treadmill input frequency and the running speed.

[0042] As a preferred technical solution, the visual feedback module is used to provide the patient with real-time gait data and training targets that contain implicit distortions, allowing the patient to modify their gait characteristics. The specific steps are as follows:

[0043] The stride length SL of the hemiplegic side during a gait cycle paretic Compared with the distorted non-hemiplegic side step length and the average of the two The visual bars are displayed on the screen, and the patient adjusts their gait according to the current gait and the target gait.

[0044] The distorted non-hemiplegic step length This refers to gradually decreasing the step length SL on the non-hemiplegic side using a decreasing amount when the controller module output increases in difficulty, i.e., ΔD > 0. non-paretic Distortion is applied.

[0045] As a preferred technical solution, the target symmetry coefficient is the hemiplegic side step length SL. paretic Compared with the distorted non-hemiplegic side step length The symmetry coefficient of the target step size when they are equal.

[0046] In summary, compared with the prior art, the effective effects of the technical solution provided in this application include at least the following:

[0047] The kinematic and dynamic information of the lower limbs was obtained by measuring the dual-track treadmill with split belts. The error and variation of the patient's own symmetry coefficient and the target symmetry coefficient were fully considered. Nonlinear factors were added to enhance the expressive power of the parameters. The fuzzy logic controller has strong robustness in controlling the training difficulty. The speed of the two tracks is adaptively adjusted separately to effectively improve the symmetry of the patient's gait. The training target is provided through visual feedback to make the training more efficient. The training target is distorted to increase implicit motor memory, so that the training effect is retained to a greater extent and a better gait rehabilitation effect is achieved. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a schematic diagram of the structural framework of an adaptive adjustable belt speed dual-track treadmill system provided in an embodiment of this application;

[0050] Figure 2 This is a schematic flowchart of an adaptive adjustment system for the belt speed of a dual-track treadmill with split belt, provided in an embodiment of this application.

[0051] Figure 3 This is a schematic diagram of the control flow of an adaptive adjustment system for the belt speed of a dual-track treadmill with split belt, provided in an embodiment of this application.

[0052] Figure 4 This is a fuzzy set curve diagram of the fuzzy logic controller provided in the embodiments of this application;

[0053] Figure 5 This is the visual feedback paradigm of the visual feedback module provided in the embodiments of this application. Detailed Implementation

[0054] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort are within the scope of protection of the present application.

[0055] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0056] Please see Figure 1 and Figure 2 In one embodiment of this application, an adaptive adjustment system for the belt speed of a split-belt dual-track treadmill is provided, comprising: an information acquisition module 201, an information analysis module 202, a controller module 203, a speed control module 204, and a visual feedback module 205.

[0057] Its adaptive speed adjustment process is as follows:

[0058] 100. Before starting rehabilitation training, measure the patient's most comfortable walking speed on a split-belt dual-track treadmill. The method for measuring the most comfortable walking speed is as follows: both belts of the split-belt dual-track treadmill start at the same speed from the minimum speed. The patient walks on the treadmill while both belts slowly accelerate at the same acceleration. When the patient reports the current speed as their most comfortable walking speed, record this speed as v. c1 Continue accelerating until the patient reports their current speed as their fastest walking speed, then begin decelerating. Record the speed as v when the patient again reports their current speed as their most comfortable walking speed. c2 The measurement is now complete. The patient's most comfortable walking speed on the dual-track treadmill with a suture zone is calculated as follows:

[0059]

[0060] 101. The information acquisition module 201 acquires the kinematic and dynamic information of the patient's lower limb joints during the patient's gait training and inputs the kinematic and dynamic information into the information analysis module 202.

[0061] The information acquisition module 201 includes a motion signal acquisition unit and a pressure signal acquisition unit.

[0062] (1) The specific steps of the motion signal acquisition unit to acquire the kinematic information of the patient's lower limb joints are as follows:

[0063] The motion signal acquisition unit is specifically an optical motion capture system, which consists of components such as a high-speed infrared data acquisition camera and a PoE switch.

[0064] In this embodiment, the kinematic information refers to the position information of 5 marker points on each side of the patient's lower limbs during walking, which can be used to calculate kinematic parameters such as the distance, angle, speed, and acceleration of each joint of the lower limbs.

[0065] (2) The specific steps for the pressure signal acquisition unit to acquire the dynamic information of the patient's lower limb joints are as follows:

[0066] The pressure signal acquisition unit is specifically a three-dimensional pressure sensor, with a total of 8 three-dimensional pressure sensors placed under each running belt of the split-belt dual-track treadmill.

[0067] In this embodiment, the dynamic information refers to the three-dimensional reaction force between the patient and the ground generated during walking. The three-dimensional pressure sensor can obtain the three-dimensional reaction force between the patient and the ground in real time by taking advantage of the characteristic that the force changes in all three directions during gait.

[0068] 102. The information analysis module 202, based on the kinematic and dynamic information acquired in real time by the information acquisition module 201, inputs the real-time kinematic information into the symmetry assessment model, analyzes the change in error between the patient's real-time symmetry coefficient and the target symmetry coefficient, and inputs the obtained average mapped error and the average change in error into the controller module 203. Specifically, this includes the following steps:

[0069] S1: After removing baseline values ​​and low-pass filtering from the real-time input dynamic information, the patient's gait events are detected and gait cycles are divided using a threshold method;

[0070] S2: Calculate the patient's symmetry coefficient within the current gait cycle using real-time input kinematic information;

[0071] S3: Compare the real-time symmetry coefficient with the target symmetry coefficient to obtain the error between the two;

[0072] S4: Compare the error of the current gait cycle with the error of the previous gait cycle to obtain the current error change;

[0073] S5: Adjust the error through nonlinear mapping.

[0074] S6: Calculate the average value of the mapping error and the error change of the patient in multiple gait cycles during the most recent training phase, and obtain the average mapping error and the average error change of the patient during this training phase.

[0075] In step S1, the method for detecting gait events based on dynamic information using a threshold approach is as follows: the first frame is when the resultant force of the vertical ground reaction force of the three-dimensional pressure sensor under any running belt of the split-belt dual-track treadmill is greater than 20N, indicating heel strike; the first frame is when the resultant force is less than 20N, indicating toe lift-off. Left toe lift-off is taken as the start of each gait cycle.

[0076] In step S2, the symmetry coefficient can typically be calculated from the spatiotemporal parameters and dynamic parameters of the gait, and can be the stride length symmetry coefficient, stride width symmetry coefficient, support time symmetry coefficient, etc., which can be calculated by the following equation:

[0077] ratio = A paretic / A non-paretic

[0078] SI = [(A paretic -A non-paretic ) / 0.5(A paretic +A non-paretic )]×100%

[0079] GA = |100×[ln(A)] paretic / Anon-paretic )]|

[0080] SA=[(45°-arctan(A paretic / A non-paretic [))×100%] / 90

[0081] Specifically, the equations in the above formulas can all be used to calculate the symmetry coefficients, where A paretic Data representing the affected side of the patient, A non-paretic This represents data from the non-hemiplegic side of the patient. This data can be kinematic information obtained by the information acquisition module 201, such as stride length, swing phase time, and stance phase time, or dynamic information obtained by the information acquisition module 201, such as forward driving force and gait mechanical pressure center. In this embodiment, the symmetry coefficient is the stride length symmetry coefficient, which is obtained by comparing the difference and mean of the stride lengths on both sides in each gait cycle of the patient.

[0082] The stride length for each gait cycle is calculated based on the forward distance between the two ankle joints when the patient's heel strikes the ground during training.

[0083] Please see Figure 3 The information acquisition module 201 will acquire the hemiplegic side step length SL in real time. paretic Stride length SL on the non-hemiplegic side non-paretic Input the symmetry assessment model to obtain the real-time step length symmetry coefficient of the patient's lower limbs. The specific calculation formula is as follows:

[0084]

[0085] Among them, SLA real-time It is the stride length symmetry coefficient of the patient's current gait cycle, SL paretic and SL non-paretic These represent the stride length on the hemiplegic side and the non-hemiplegic side, respectively, during the patient's current gait cycle.

[0086] In step S3, the error between the real-time symmetry coefficient and the target symmetry coefficient represents the difference between the symmetry of the patient's current gait cycle and the symmetry of the training target, and its calculation formula is as follows:

[0087] E = SLA target -SLA real-time

[0088] Where E is the error between the real-time step length symmetry coefficient and the target step length symmetry coefficient of the patient's current gait cycle, SLA target It is the target step length symmetry coefficient of the patient's current gait cycle.

[0089] In step S4, the error change represents the change in the gap between the patient and the training target during the training process, and its calculation formula is as follows:

[0090] ΔE=EE'

[0091] Where ΔE is the error change in the patient's current gait cycle, and E' is the error between the real-time step length symmetry coefficient and the target step length symmetry coefficient of the patient's previous gait cycle.

[0092] In step S5, the nonlinear mapping of the error refers to adjusting the error through a nonlinear transformation to obtain the mapped error. When the patient adapts to the current training difficulty, the error between the real-time step size symmetry coefficient and the target step size symmetry coefficient is small. However, when the patient has difficulty adapting to the current training difficulty, the error between the real-time step size symmetry coefficient and the target step size symmetry coefficient is large. Nonlinear mapping further reduces the smaller error and further increases the larger error, thus enhancing the expressive power of this parameter. The calculation formula for this nonlinear mapping is:

[0093]

[0094] Among them, E r is the mapping error of the patient's current gait cycle, m is the adjustment parameter for the steepness of the curve in the nonlinear transformation, and b is the adjustment parameter for the mapping range of E in the nonlinear transformation.

[0095] In step S6, since the dual-track treadmill training is a slow form of exercise adaptation training, analyzing the average parameters across multiple gait cycles is a more effective way to reflect the patient's adaptation to the current difficulty level compared to analyzing parameters from a single gait cycle. The calculation formula is as follows:

[0096]

[0097]

[0098] in, and These are the error change and the average value of the mapped error during the N gait cycles in the most recent training phase of the patient's walking training process, respectively.

[0099] 103. The controller module 203, based on the average mapped error and average error change of multiple gait cycles of the patient over a period of time calculated by the information analysis module 202, analyzes the change in training difficulty required by the patient in the next period of time through a fuzzy logic controller. The control process of the fuzzy logic controller has three steps: fuzzification, fuzzy inference, and defuzzification, specifically:

[0100] The fuzzification refers to converting the averaged mapping error and the average change in error into linguistic variables, and obtaining the membership degree of the averaged mapping error and the average change in error to the fuzzy set based on the membership function.

[0101] More specifically, in the fuzzification process, the average mapped error and the average change in error are first multiplied by a quantization factor, transforming the average mapped error and the average change in error from actual variables into the universe values ​​of the membership function in this embodiment. The calculation formula is as follows:

[0102]

[0103]

[0104] in, It is the universe of discourse value of the error after average mapping. q1 and q2 are the universe of discourse values ​​for the average change in error, respectively, and the quantization factors for the error and the average change in error after average mapping. The quantization factor is the ratio of the actual variable to the universe of discourse value, and needs to be set according to the changes in the patient's error and the average change in error after average mapping, and the universe of discourse. Then, the universe of discourse values ​​for the error and the average change in error after average mapping are input into the fuzzy set and converted into fuzzy language. The fuzzy set of the fuzzy logic controller consists of membership functions, which can be trapezoidal membership functions, triangular membership functions, Gaussian membership functions, etc.

[0105] Please see Figure 4 In this embodiment, the membership functions are triangular membership functions. Both the input and output fuzzy sets are composed of five triangular membership functions: negative large (NB), negative small (NS), zero (Z), positive small (PS), and positive large (PB). During fuzzification, the universe values ​​of the averaged error and the average change of error are mapped to two membership functions, respectively, resulting in two membership degrees for each.

[0106] The fuzzy reasoning refers to deriving fuzzy conclusions based on the membership degrees obtained by mapping two inputs through fuzzy sets and the preset fuzzy rules.

[0107] More specifically, referring to the fuzzy rule table of the fuzzy logic controller in Table 1, the fuzzy rule table is predefined based on experience, and the membership degree of the output is derived from the membership degree of the two universe values ​​using the if A and B than C rule.

[0108] Table 1

[0109]

[0110] Defuzzification refers to using defuzzification methods to map fuzzy conclusions to specific non-fuzzy outputs, i.e., the change in training difficulty currently required by the patient.

[0111] More specifically, this embodiment uses the region center method as a defuzzification method to solve for the nonfuzzy output of the fuzzy logic controller. This method uses the center of the area enclosed by the membership function curve of the output fuzzy set and the horizontal axis as the controller output. The calculation formula is as follows:

[0112]

[0113] Among them, u out It is the output of the fuzzy logic controller, c(i) is the center value of the area enclosed by the membership function curve in the fuzzy set and the horizontal axis, and u c (i) is the membership degree corresponding to c(i); then the output of the fuzzy logic controller is converted into the change in training difficulty required by the patient, and the calculation formula is:

[0114] ΔD=u out ×q3

[0115] Where ΔD is the change in training difficulty required by the patient, and q3 is the transformation coefficient between the output of the fuzzy logic controller and the change in training difficulty required by the patient.

[0116] 104. The speed control module 204 adjusts the input frequency of the two treadmill belts based on the change in training difficulty obtained by the controller module 203, thereby adjusting the running speed of the two treadmill belts separately.

[0117] More specifically, the speed control module 204 adjusts the input frequency f of the dual-track treadmill according to the change in training difficulty ΔD. paretic with f non-paretic and f paretic with f non-paretic The ratio is limited to 1:3 to prevent excessively high speed ratios from causing lower limb pain in patients. Finally, f paretic with f non-paretic The data is transmitted to the motor controller to control the motor speed of the two tracks and adjust the running speed v of the two tracks. paretic With v non-paretic The calculation formula is:

[0118]

[0119] f paretic =f paretic '-rΔD

[0120] f non-paretic =f non-paretic '+rΔD

[0121] v paretic =sf paretic

[0122] v non-paretic =sf non-paretic

[0123] Among them, f paretic with f non-paretic These are the input frequencies of the tracks containing the hemiplegic and non-hemiplegic sides during the patient's next training phase, f paretic 'and f non-paretic 'These are the input frequencies of the hemiplegic and non-hemiplegic sides on the track during the patient's previous training phase, v paretic With v non-paretic These are the running speeds of the treadmill on the hemiplegic and non-hemiplegic sides of the patient's current gait cycle, respectively; r is the transformation coefficient between the change in training difficulty ΔD required by the patient and the input frequency; and s is the transformation coefficient between the treadmill input frequency and the running speed.

[0124] 105. The visual feedback module 205 provides the patient with real-time gait data and training targets with implicit distortions in the most recent gait cycle based on the training difficulty change obtained by the controller module 203, so that the patient can change their own gait characteristics and train efficiently.

[0125] For more details, please see Figure 5 The visual feedback module first acquires the stride length SL of the hemiplegic side in the most recent gait cycle. paretic Stride length SL on the non-hemiplegic side non-paretic Then, the non-hemiplegic side step length SL non-paretic Distortion is applied, and the step length SL on the hemiplegic side is calculated. paretic Compared with the distorted non-hemiplegic side step length The average of the values ​​is the average step length SL. Finally, the step length SL of the hemiplegic side is calculated. paretic ( Figure 5 (Solid line bar chart on the left) and the distorted step length on the non-hemiplegic side ( Figure 5 (The solid line bar chart on the right) and the average step size of both. The steps are displayed on the screen as visual bars, which mislead patients into thinking that the displayed step length is the step length of their most recent gait cycle, and then they adjust their gait according to the current step length and the target step length.

[0126] Step length on the non-hemiplegic side Distortion refers to gradually decreasing the step length SL on the non-hemiplegic side by a decreasing amount when the controller module output increases in difficulty (ΔD > 0). non-paretic The calculation formula is:

[0127]

[0128] Where d is the decrease in distortion per step, and n is the number of distortions. maxThis represents the maximum number of distortions, limiting the maximum degree of distortion. It is achieved by providing a distorted step length on the non-hemiplegic side. As a goal, patients will subconsciously believe that the step length on the non-hemiplegic side needs to be increased to achieve symmetry, thereby implicitly increasing the training volume on the hemiplegic side and improving the rehabilitation effect of walking training.

[0129] In this embodiment, the target step length symmetry coefficient is the step length SL on the hemiplegic side. paretic Compared with the distorted non-hemiplegic side step length The step size symmetry when they are equal is calculated using the following formula:

[0130]

[0131] Among them, SLA target It is the target step length symmetry coefficient of the patient's current gait cycle; by combining the distorted target step length symmetry coefficient as the patient's training target, and by adjusting the current training difficulty according to the gap between the patient's current gait characteristics and the target through a fuzzy logic controller, the patient's walking training will be more efficient and achieve better rehabilitation results.

[0132] In summary, this application provides an adaptive adjustment system for the belt speed of a split-belt dual-track treadmill. It utilizes the split-belt dual-track treadmill to measure the kinematic and dynamic information of the lower limbs, fully considering the error and variation between the patient's own symmetry coefficient and the target symmetry coefficient. The addition of nonlinear factors enhances the expressive power of the parameters. The fuzzy logic controller exhibits strong robustness in controlling training difficulty. The system adaptively adjusts the speed of the two belts separately, effectively improving the symmetry of the patient's gait. Visual feedback provides training targets, increasing training efficiency. Distortion of the training targets enhances implicit motor memory, allowing for greater retention of training effects and achieving better gait rehabilitation results.

[0133] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0134] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0135] The above embodiments are preferred embodiments of this application, but the implementation of this application is not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of this application shall be considered equivalent substitutions and shall be included within the protection scope of this application.

Claims

1. An adaptive adjustment system for the belt speed of a dual-track treadmill with split belts, characterized in that, include: Information acquisition module, information analysis module, controller module, speed control module, and visual feedback module; The information acquisition module is used to acquire kinematic and dynamic information of the patient's lower limb joints in real time during the patient's walking training. The information analysis module is used to calculate the average mapping error and the average change of error between the real-time symmetry coefficient of the patient's lower limb and the target symmetry coefficient based on the kinematic and dynamic information acquired in real time by the information acquisition module. The controller module is a fuzzy logic controller, which is used to deduce the appropriate training difficulty change for the patient based on the average mapping error and the average change of error using fuzzy rules. After each training period, the fuzzy logic controller obtains the patient's parameter changes and calculates the required training difficulty change for the patient in the next period according to fuzzy rules. The speed control module is used to adjust the input frequency of the treadmill motor according to the change in training difficulty, and to individually adjust the running speed of the two treadmill belts. Specifically, the speed control module is used to adjust the motor input frequency of the two belts of the split-belt dual-track treadmill according to the change in training difficulty required by the patient, and to limit the ratio of the motor input frequencies of the two belts to 1:3, thereby individually adjusting the running speed of the two treadmill belts. The specific calculation formula is as follows: in, and These are the motor input frequencies of the tracks located on the hemiplegic and non-hemiplegic sides during the patient's next training phase. and These are the motor input frequencies of the tracks located on the hemiplegic and non-hemiplegic sides during the patient's previous training phase. and These are the running speeds of the tracks on the hemiplegic and non-hemiplegic sides, respectively, during the patient's current gait cycle. This is the amount of change in training difficulty currently required by the patient. Transformation coefficients relative to the input frequency, It is the conversion coefficient between the treadmill's input frequency and its running speed; The visual feedback module is used to provide the patient with real-time gait data containing implicit distortions and training targets, allowing the patient to modify their gait characteristics. Specifically, the visual feedback module provides the patient with real-time gait data containing implicit distortions and training targets, allowing the patient to modify their gait characteristics, and the specific steps are as follows: The stride length of the hemiplegic side during a gait cycle Compared with the distorted non-hemiplegic side step length and the average of the two The visual bars are displayed on the screen, and the patient adjusts their gait according to the current gait and the target gait. The distorted non-hemiplegic step length This refers to the increased difficulty when the controller module outputs information. At that time, the step length on the non-hemiplegic side was gradually reduced using a decreasing amount. Distortion is applied.

2. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 1, characterized in that, The information acquisition module also includes a motion signal acquisition unit and a pressure signal acquisition unit; The motion signal acquisition unit is used to collect kinematic information of each joint of the patient's lower limbs during movement in real time; The pressure signal acquisition unit is used to collect the three-dimensional reaction force between the patient and the track during walking in real time.

3. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 1, characterized in that, The information analysis module is specifically used to input the real-time acquired kinematic and dynamic information into the symmetry assessment model during the patient's walking process, and analyze the changes in the error parameters between the patient's real-time symmetry coefficient and the target symmetry coefficient, including the following steps: After removing baseline values ​​and low-pass filtering from the real-time acquired dynamic information, gait events are detected and gait cycles are divided using a threshold method. The patient's symmetry coefficient within the current gait cycle is calculated using real-time acquired kinematic information. By comparing the real-time symmetry coefficient with the target symmetry coefficient, the error between the two in the current gait cycle is obtained; Compare the error of the current gait cycle with the error of the previous gait cycle to obtain the current error change. The error of the current gait cycle is adjusted by nonlinear mapping to obtain the mapped error; The average value of the mapping error and the average change in error of the patient in multiple gait cycles within a training phase are calculated to obtain the average mapping error and the average change in error of the patient within that training phase.

4. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 3, characterized in that, The symmetry coefficient is the step-size symmetry coefficient, and its calculation formula is as follows: in, It is the stride length symmetry coefficient of the patient's current gait cycle. and These represent the stride length on the hemiplegic side and the stride length on the non-hemiplegic side during the patient's current gait cycle.

5. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 3, characterized in that, The error of the current gait cycle is adjusted through nonlinear mapping to obtain the mapped error. The specific calculation formula is as follows: in, It is the error after mapping the patient's current gait cycle. It is the error between the real-time step length symmetry coefficient of the patient's current gait cycle and the target step length symmetry coefficient. It is an adjustment parameter for the steepness of the curve in nonlinear transformation. yes Adjustment parameters for the mapping range in nonlinear transformations.

6. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 1, characterized in that, The controller module is a fuzzy logic controller, and the control process includes fuzzification, fuzzy inference, and defuzzification. The fuzzification refers to converting the average mapping error and the average change in error into linguistic variables, and obtaining the membership degree of the average mapping error and the average change in error to the fuzzy set according to the membership function. The fuzzy reasoning refers to deriving fuzzy conclusions based on the input membership degree and preset fuzzy rules; Defuzzification refers to using defuzzification methods to map fuzzy conclusions to specific non-fuzzy outputs, i.e., the change in training difficulty currently required by the patient.

7. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 6, characterized in that, The membership function is a trigonometric membership function; The defuzzification method is the region center method, which uses the center of the area enclosed by the membership function curve of the output fuzzy set and the horizontal axis as the output of the fuzzy logic controller, and then converts the output of the fuzzy logic controller into the change in training difficulty required by the patient.

8. The adaptive adjustment system for the belt speed of a dual-track treadmill with split belts according to claim 1, characterized in that, The target symmetry coefficient is the step length on the hemiplegic side. Compared with the distorted non-hemiplegic side step length The symmetry coefficient of the target step size when they are equal.

Images