A closed-loop control method and system for lower limb prosthesis

Abstract

The present invention discloses a closed-loop control method and system for a lower limb prosthesis. The closed-loop control method comprises: respectively collecting intrinsic myoelectric signals of the residual limb lower limb muscles and plantar pressure signals of the intact leg; identifying the current walking state based on the intrinsic myoelectric signals; issuing a control command to the prosthesis based on the current walking state to drive the prosthesis to a specified position; inputting the plantar pressure signals of the intact leg into a gait recognition model to identify and output the current gait information; querying a large database of gait plantar pressure information based on the current gait information to obtain a predicted value of the pressure distribution corresponding to the plantar of the prosthesis; collecting the plantar pressure signals of the prosthetic lower limb, comparing the plantar pressure signals of the prosthetic lower limb with the predicted value of the pressure distribution corresponding to the plantar of the prosthesis to obtain a deviation signal; and performing feedback control on the prosthesis movement based on the deviation signal. The closed-loop control of the lower limb prosthesis is achieved with automatic deviation correction capability, greatly improving the anti-interference ability and precision of the prosthesis control.

Description

Technical Field

[0001] The present invention relates to the technical field of prostheses, and in particular to a closed-loop control method and system for lower limb prostheses. Background Art

[0002] Every year, a large number of people around the world undergo lower limb amputations to save their lives due to work-related injuries, illness, traffic accidents, and war. Amputations not only cause physical disability but also severe psychological trauma, severely impacting patients' work and personal lives. According to a national survey of people with disabilities, the number of people with physical disabilities in my country is approximately 20 million, yet less than 10% of these individuals receive basic rehabilitation services, placing a significant burden on their families and society.

[0003] Currently, amputees typically use prosthetic limbs to restore walking function. Prosthetic technology is developing towards advanced technologies that are more precise, comfortable, and tailored to individual needs. Most prosthetic limbs utilize open-loop myoelectric control using surface electrodes. These sensors are placed on the body surface, making them susceptible to interference, motion artifacts, and environmental influences. Furthermore, this open-loop system cannot guarantee accurate movement of the prosthesis in response to control signals, making it susceptible to interference and resulting in low accuracy.

[0004] Therefore, based on the existing prosthetic technology, how to provide a closed-loop control method and system for lower limb prostheses to solve the problems of weak anti-interference, easy electrode detachment, discomfort and allergy induction mediated by traditional prosthetic surface electrodes has become an urgent problem that needs to be solved by technical personnel in this field. Summary of the Invention

[0005] In view of the above problems, the present invention proposes a closed-loop control method and system for lower limb prostheses that at least solves some of the above technical problems. The method and system have the automatic correction capability of closed-loop control and can greatly improve the anti-interference ability and accuracy of the prosthetic control system.

[0006] An embodiment of the present invention provides a closed-loop control method for a lower limb prosthesis, comprising:

[0007] The method includes collecting the intrinsic myoelectric signals of the lower limb muscles of the residual limb and the plantar pressure signals of the good leg respectively; identifying the current walking state based on the intrinsic myoelectric signals; issuing control instructions to the prosthesis based on the current walking state to drive the prosthesis to move to a specified position; inputting the plantar pressure signals of the good leg into a gait recognition model to identify and output the current gait information;

[0008] According to the current gait information, query the gait plantar pressure information database to obtain the pressure distribution prediction value corresponding to the prosthetic foot; the gait plantar pressure information database includes: plantar pressure distribution data and gait information of the left and right legs during normal walking of the human body;

[0009] The pressure signal of the plantar of the prosthetic lower limb is collected, and the pressure signal of the plantar of the prosthetic lower limb is compared with the predicted value of the pressure distribution corresponding to the plantar of the prosthetic limb to obtain a deviation signal; and the prosthetic movement is feedback controlled according to the deviation signal.

[0010] Furthermore, the step of identifying the current walking state based on the internal electromyographic signal includes the following steps:

[0011] S11, filtering and amplifying the internal electromyographic signal;

[0012] S12, converting the filtered and amplified internal electromyographic signal into a digital signal through an AD chip;

[0013] S13, analyzing the digital signal using a fixed sample entropy peak-valley threshold algorithm, extracting peaks and troughs of the internal electromyographic signal sample entropy, and obtaining gait phase information;

[0014] S14: querying a large gait information database based on the gait phase information to identify the current walking state.

[0015] Furthermore, the establishment of the gait recognition model includes the following steps:

[0016] S21, selecting plantar pressure distribution data and gait information of the left and right legs of a human body during normal walking as input and output data of a pre-built neural network model; the plantar pressure distribution data and gait information of the left and right legs are obtained from the gait plantar pressure information database;

[0017] S22, iteratively training the neural network model to find the optimal parameters based on the cost equation error of the model;

[0018] S23. When the cost equation error is minimized, optimal parameters are obtained to obtain a constructed gait recognition model.

[0019] Furthermore, the feedback control of the prosthetic limb movement according to the deviation signal includes: processing the deviation signal through a PID algorithm, outputting a corresponding control signal, and performing feedback control on the prosthetic limb movement.

[0020] The embodiment of the present invention further provides a lower limb prosthesis closed-loop control system, which is applicable to any of the above lower limb prosthesis closed-loop control methods, comprising: a host computer, an electrode wire, a controller, and a pressure sensor;

[0021] The electrode wire is used to collect the internal electromyographic signals of the residual limb lower limb muscles;

[0022] The pressure sensor is used to collect the pressure signal of the good leg plantar and the pressure signal of the prosthetic lower limb plantar, and transmit them to the controller;

[0023] The controller is connected to the host computer, the electrode wire, the pressure sensor and the prosthesis respectively, and is used to wirelessly transmit the internal myoelectric signal to the host computer and issue control instructions to the prosthesis according to the current walking state identified by the host computer;

[0024] The host computer is used to identify and output the current walking state according to the internal electromyographic signal.

[0025] Furthermore, the controller is further configured to wirelessly transmit the plantar pressure signal of the good leg to the host computer, and compare the predicted pressure distribution corresponding to the prosthetic foot identified by the host computer with the plantar pressure signal of the prosthetic lower limb to generate a deviation signal and perform feedback control on the prosthetic movement;

[0026] The host computer is further configured to identify and output a predicted pressure distribution value corresponding to the sole of the prosthetic limb based on the sole pressure signal of the good leg.

[0027] Furthermore, the pressure sensor adopts a flexible thin film pressure sensor ZNX-01.

[0028] The beneficial effects of the above technical solutions provided by the embodiments of the present invention include at least:

[0029] An embodiment of the present invention provides a closed-loop control method for a lower limb prosthesis, comprising: separately collecting intrinsic myoelectric signals of the residual limb lower limb muscles and plantar pressure signals of the intact leg; identifying the current walking state based on the intrinsic myoelectric signals; issuing a control command to the prosthesis based on the current walking state, driving the prosthesis to move to a specified position; inputting the plantar pressure signals of the intact leg into a gait recognition model to identify and output current gait information; querying a large database of gait plantar pressure information based on the current gait information to obtain a predicted value of the pressure distribution corresponding to the plantar of the prosthesis; collecting the plantar pressure signals of the prosthetic lower limb, comparing the plantar pressure signals of the prosthetic lower limb with the predicted value of the pressure distribution corresponding to the plantar of the prosthesis to obtain a deviation signal; and performing feedback control of the prosthesis movement based on the deviation signal. This method achieves closed-loop control of the lower limb prosthesis with automatic deviation correction capability, greatly improving the anti-interference ability and accuracy of prosthesis control.

[0030] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purposes and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0031] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0033] Figure 1 A flow chart of a closed-loop control method for a lower limb prosthesis provided by an embodiment of the present invention;

[0034] Figure 2 A flowchart of myoelectric signal processing and analysis in lower limb muscles provided by an embodiment of the present invention;

[0035] Figure 3 A flowchart of good leg plantar pressure signal processing and analysis provided by an embodiment of the present invention;

[0036] Figure 4 A schematic diagram of the arrangement and distribution of plantar pressure sensors provided in an embodiment of the present invention;

[0037] Figure 5 Schematic diagram of a closed-loop control system for a lower limb prosthesis provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0039] The embodiment of the present invention provides a closed-loop control method for lower limb prosthesis, referring to Figure 1 As shown, including:

[0040] The system collects the intrinsic myoelectric signals of the residual limb lower limb muscles and the plantar pressure signals of the good leg respectively; identifies the current walking state based on the intrinsic myoelectric signals; issues control instructions to the prosthesis based on the current walking state, driving the prosthesis to move to the specified position; inputs the plantar pressure signals of the good leg into the gait recognition model, and identifies and outputs the current gait information;

[0041] Based on the current gait information, the gait plantar pressure information database is queried to obtain the corresponding pressure distribution prediction value of the prosthetic foot. The gait plantar pressure information database contains: the plantar pressure distribution data of the left and right legs during normal walking (corresponding to the pressure distribution prediction value) and gait information;

[0042] The pressure signal of the plantar of the prosthetic lower limb is collected, and the pressure signal of the plantar of the prosthetic lower limb is compared with the predicted value of the pressure distribution corresponding to the plantar of the prosthetic limb to obtain a deviation signal; the prosthetic movement is feedback controlled according to the deviation signal.

[0043] The closed-loop control method for lower limb prostheses provided in this embodiment combines forward control with feedback control to jointly drive the prosthesis to move to the specified position that the user wants to reach, with strong anti-interference ability and high precision. By collecting internal electromyographic signals, the problems of susceptibility to interference, motion artifacts and environmental influences, as well as the easy detachment of electrodes, discomfort and allergic reactions caused by the use of surface electrodes are avoided. The collected electromyographic signals are used for gait recognition and control of prosthetic movements. At the same time, the plantar pressure signals of the legs and prostheses are also collected. Combined with the gait recognition model based on the plantar pressure signals, closed-loop control of the lower limb prostheses is achieved, avoiding the shortcomings of open-loop control such as low precision and lack of automatic deviation correction capability, thereby greatly improving the anti-interference ability and precision of the prosthetic control system.

[0044] The following is a detailed description of the implementation process of this method:

[0045] This embodiment provides a closed-loop control method for lower limb prostheses (applicable to single-leg amputation). The collected internal electromyographic signals are used for gait recognition, and control instructions are issued to the intelligent prosthesis based on the recognized gait to directly drive the prosthetic movement. During the movement of the prosthesis, the plantar pressure signals of the good leg and the lower end of the prosthesis are also collected in real time. The plantar pressure signals of the good leg are used as the input of the gait recognition algorithm (gait recognition model) based on the plantar pressure signals, and the current gait information can be obtained (the two legs together constitute the current gait information, and the gait information reflects the corresponding relationship between the plantar pressure distribution of the two legs). Then, based on the output current gait information, the gait plantar pressure information database is queried to predict the corresponding pressure distribution of the prosthesis at this time. The plantar pressure signal of the prosthesis detected at this time is then compared with the pressure distribution corresponding to the plantar pressure of the prosthesis predicted above to obtain a deviation signal. The deviation signal is then converted into a corresponding control signal through the PID algorithm, thereby performing feedback control on the prosthetic movement and realizing closed-loop control of the lower limb prosthesis. After directly driving the prosthetic limb to move, the plantar pressure signal of the prosthetic limb is continuously collected as a feedback signal to continuously eliminate deviations.

[0046] Specifically, the system collects the internal electromyographic signals of the lower limb muscles of the residual limb to identify the current walking state; based on the current walking state, it sends control instructions to the prosthesis to drive the prosthesis to move to the specified position.

[0047] Implanting electrode wires into the muscles of the lower limb of the residual limb enables the acquisition of intrinsic myoelectric signals. The collected signals are then filtered and amplified, converted into digital signals through an AD chip, transmitted to a controller, and analyzed using a specific algorithm. For example, a fixed sample entropy peak-valley threshold algorithm can be used. This algorithm extracts the peaks and troughs of the intrinsic myoelectric signal sample entropy to obtain gait phase information. This is then compared with the information of each stage in the gait cycle in the established gait information database, and the walking state corresponding to each stage information (support phase or swing phase) is queried, thereby realizing the identification of the current walking state (identifying whether it is the support phase or the swing phase at this time). Among them, during the normal walking process of the human body, the foot is constantly repeating two states, namely the support phase and the swing phase. Here, a gait cycle includes a support phase and a swing phase.

[0048] The gait information database includes: a peak-to-trough waveform of the entropy of the electrical signal samples for a complete gait cycle; and the gaits corresponding to different stages in a gait cycle. The gait is the corresponding position of the two feet at that moment.

[0049] By comparing the peaks and troughs of the extracted internal electromyographic signal sample entropy with a gait cycle, we can determine which stage of the cycle it is in, and then by querying the gait information database, we can get the current gait (stance phase or swing phase).

[0050] The controller sends corresponding control instructions to the prosthesis based on the walking state it recognizes, driving the prosthesis to move. Under this control instruction, the prosthesis can be moved directly to the designated position. The deviation in this process is very small, which greatly reduces the subsequent adjustment time, has small fluctuations, and enhances control stability. Figure 2 shown.

[0051] Specifically, the leg and foot pressure signals are collected and input into the gait recognition model to identify and output the current gait information; based on the current gait information, the gait and foot pressure information database is queried to obtain the corresponding pressure distribution prediction value of the prosthetic foot. Specifically, it includes:

[0052] Optionally, taking into account factors such as comfort, price, performance, and scope of application, the pressure sensor selected is the flexible film pressure sensor ZNX-01. Other types of pressure sensors can also be selected according to actual needs, and this embodiment does not limit them. The collected good leg foot pressure signal is used as the input of the gait recognition model to obtain the current gait information, and then the gait plantar pressure information database is queried to obtain the corresponding pressure distribution prediction value of the prosthetic foot at this time. For the specific analysis process, refer to Figure 3 shown.

[0053] Reference Figure 4Figure 2 shows a schematic diagram of the plantar pressure sensor placement. The primary load-bearing areas of the plantar foot are located near the distal phalanges, sesamoids, proximal phalanges, metatarsophalangeal radiculone, and cuboid bones. Therefore, pressure sensors can be placed in these areas to collect pressure signals. Pressure sensors are implanted in the insoles to collect plantar pressure signals. The collected pressure signals include both the intact leg and the prosthetic lower limb.

[0054] During normal walking, the foot repeatedly repeats two phases: the stance phase and the swing phase. The stance phase is the period from when the foot contacts the ground to when it leaves it, specifically starting with heel strike and ending with toe lift-off. During this phase, the supporting leg maintains contact with the ground, the supporting person's center of gravity gradually shifts forward, and the angle between the plantar plane and the ground constantly changes. This phase accounts for a significant portion of the entire gait cycle. The swing phase is when the foot leaves the ground and remains suspended in the air, starting with toe lift-off and ending with heel strike. During this phase, the foot remains suspended in the air and swings from back to front in an arc-shaped motion. The swing phase accounts for a relatively small portion of the entire gait cycle. During the stance phase, the human foot's motion characteristics (gait cycle) can be broken down into: heel strike (HS), toe-on, midstance, heel-off, and toe-off.

[0055] During normal walking, the plantar pressure distributions of the left and right legs correspond to each other. This correspondence can be established by testing healthy volunteers of different ages and genders (without foot diseases or physical disabilities) while walking. The distribution of plantar pressure signals of the test subjects' left and right legs is collected, and a large database of gait plantar pressure information is ultimately established. This database is then used in the subsequent gait recognition classification algorithm (gait recognition model) to optimize model parameters and ultimately establish a gait recognition model with optimal parameters based on the plantar pressure signals. It is not possible to directly query the large database of gait plantar pressure information based on the collected plantar pressure signal of the good leg because this database cannot fully encompass all possibilities. Therefore, when establishing a gait recognition model, the current gait information (each gait information reflects the corresponding plantar pressure distribution of the two legs) is obtained, and then the large database of gait plantar pressure information is subsequently queried to obtain the corresponding pressure distribution value of the prosthetic foot. In other words, a continuous functional relationship is obtained from limited data, meaning that any input can be predicted and has an output.

[0056] Among them, the construction of the gait recognition model based on the plantar pressure signal can adopt a machine learning algorithm, such as dynamic time warping, support vector machine algorithm, etc., or other algorithms such as neural network algorithm, which are not limited in this embodiment. Taking the support vector machine algorithm as an example, first a neural network model is given, and an initial value is given for the model parameters. Secondly, in order to facilitate the algorithm to perform gait classification and recognition, the data can be feature extracted according to the five gait stages of human foot movement - heel strike (HS, heel-strike), toe follow-up moment (toe-on), full palm landing moment (mid-stance), heel off (heel-off), toe off (toe-off), and data normalization is performed on the data to improve the network convergence speed, and the parameters of the model are continuously iterated to modify the cost equation (error) of the model. Finally, the optimal model parameters are obtained, that is, the corresponding gait recognition model with the optimal parameters based on the plantar pressure signal is obtained. The specific process is:

[0057] ① Input database data. Input known input and output data from the gait plantar pressure information database, including the plantar pressure distribution data of the left and right legs and gait information during normal walking, to train and optimize model parameters.

[0058] ②Perform data normalization to improve network convergence speed. Normalize the input and output data, calculate the 2-norm of the input and output data respectively, and then divide each data element in the input and output data by the corresponding 2-norm, which improves the convergence speed and reduces the number of model iterations.

[0059] ③ Parameter optimization and grid number search begin. Iterate the model training to find the optimal parameters.

[0060] ④ Obtain the optimal parameters to establish a prediction model. Obtain the optimal model parameters and establish an optimal parameter gait recognition model. Input the plantar pressure signal of the good leg and output the current gait information.

[0061] ⑤ Prediction data input: Input the plantar pressure signal of the good leg to be predicted to the trained gait recognition model.

[0062] ⑥ The classification result is the current gait. The positive judgment outputs the current gait information.

[0063] The model then uses the current gait information output by the model to query a large database of gait plantar pressure information to obtain the corresponding prosthetic plantar pressure signal (pressure distribution prediction value). Each gait information in the database corresponds to multiple plantar pressure signals. The average of these plantar pressure signals is the plantar pressure distribution corresponding to the gait information. Therefore, the predicted pressure distribution value corresponding to the current prosthetic plantar pressure can be obtained based on the determined gait information.

[0064] Specifically, the pressure signal of the prosthetic lower limb plantar is collected, and the pressure signal of the prosthetic lower limb plantar is compared with the predicted value of the pressure distribution corresponding to the prosthetic plantar to obtain a deviation signal; and the prosthetic movement is feedback controlled according to the deviation signal. Specifically, it includes:

[0065] The pressure signal at the lower end of the prosthetic foot (the predicted pressure distribution at the prosthetic foot) is used as feedback in the closed-loop control process. The collected pressure signal is compared with the predicted pressure distribution at the prosthetic foot, as predicted by the pressure signal at the unaffected leg. This generates a deviation signal, which is processed using a PID algorithm and ultimately outputs a corresponding control signal. This provides feedback control of the prosthetic movement, thus achieving closed-loop control of the lower limb prosthesis.

[0066] Specifically, the PID control algorithm is a control algorithm that combines the three control links of proportion, integration and differentiation. It is the most mature and widely used control algorithm in continuous systems. The essence of PID control is to perform calculations based on the functional relationship of proportion, integration and differentiation according to the input deviation value, and the calculation results are used to control the output.

[0067] The closed-loop feedback control method provided in this embodiment is applicable to discrete systems. It converts the continuous PID control into digital form, replaces the differential link with the difference, replaces the integral link with the cumulative sum, and keeps the proportional link unchanged.

[0068] This embodiment provides a closed-loop control method for a lower limb prosthesis. The method uses collected internal electromyographic signals to identify the current walking state and issues control instructions to the intelligent prosthesis based on the identified current walking state. During the prosthetic movement, the method also collects real-time pressure signals from the plantar pressure of the good leg and the lower end of the prosthesis. The plantar pressure signals of the good leg are used as input to a gait recognition model based on the plantar pressure signals to obtain current gait information. The gait plantar pressure information database is then queried based on the output current gait information to predict the pressure distribution corresponding to the plantar pressure of the prosthesis at that time. The detected plantar pressure signal of the prosthesis is then compared with the predicted pressure distribution corresponding to the plantar pressure of the prosthesis to obtain a deviation signal. The deviation signal is then converted into a corresponding control signal using a PID algorithm, thereby providing feedback control of the prosthetic movement. This method implements closed-loop control of the lower limb prosthesis, provides automatic deviation correction capabilities, and significantly improves the anti-interference ability and accuracy of the prosthetic control system.

[0069] Based on the same inventive concept, embodiments of the present invention also provide a closed-loop control system for a lower-limb prosthesis, applicable to the closed-loop control method for a lower-limb prosthesis described above. Because the principles underlying the problem solved by this system are similar to those of the aforementioned closed-loop control method for a lower-limb prosthesis, the implementation of this system can be referenced to the aforementioned method, and any repetitions will not be repeated.

[0070] A closed-loop control system for a lower limb prosthesis, comprising: a host computer (computer or mobile phone), an electrode wire, a controller, and a pressure sensor;

[0071] The electrode wire is implanted in the muscles of the residual limb lower limb to collect the intrinsic myoelectric signals of the muscles of the residual limb lower limb;

[0072] The pressure sensors are respectively arranged on the sole of the good leg and the sole of the prosthetic lower limb, for respectively collecting the pressure signal of the sole of the good leg and the sole of the prosthetic lower limb and transmitting the pressure signal to the controller;

[0073] The controller is connected to the host computer, electrode wire, pressure sensor and prosthesis respectively, and is used to wirelessly transmit the internal myoelectric signals to the host computer and issue control instructions to the prosthesis based on the current walking state recognized by the host computer. It also transmits the pressure signal of the plantar of the good leg to the host computer and compares the pressure distribution corresponding to the plantar of the prosthetic limb recognized by the host computer with the pressure signal of the plantar of the prosthetic lower limb to generate a deviation signal for feedback control of the prosthetic movement.

[0074] The host computer is used to identify and output the current walking state based on the internal electromyographic signal; and to identify and output the corresponding pressure distribution prediction value of the prosthetic foot based on the plantar pressure signal of the good leg.

[0075] This embodiment provides a closed-loop control system for lower limb prostheses (applicable to single-leg amputation). The internal electromyographic signals collected by the electrode wire are used for gait recognition. According to the recognized gait, the controller sends control instructions to the intelligent prosthesis. Figure 5 As shown, it is a schematic diagram of the closed-loop control system of the lower limb prosthesis provided by this embodiment. The control instruction is equivalent to the initial disturbance signal in the entire closed-loop control system. Under this control instruction, the prosthesis can be moved directly to the vicinity of the specified position, that is, the deviation in the closed-loop control system is very small, thereby greatly reducing the adjustment time of the control system, with small fluctuations and improved system stability. During the movement of the prosthesis, the pressure sensor also collects the plantar pressure signals of the good leg and the lower end of the prosthesis in real time. The plantar pressure signal of the good leg is used as the input of the gait recognition algorithm (gait recognition model) based on the plantar pressure signal to obtain the current gait information. Then, according to the output current gait information, the gait plantar pressure information database is queried to predict the pressure distribution corresponding to the plantar of the prosthesis at this time, which is the attached Figure 5 Then the pressure signal of the prosthetic foot detected at this time is compared with the pressure distribution corresponding to the prosthetic foot predicted by the given link to obtain the deviation signal, which is the additional Figure 5The comparison link in the closed-loop control system. Finally, the PID algorithm converts the deviation signal into a corresponding control signal, thereby providing feedback control of the prosthetic movement and achieving closed-loop control of the lower limb prosthesis by the controller. After directly driving the prosthetic movement, the plantar pressure signal of the prosthetic foot is continuously collected as feedback signal, continuously approaching the given link (i.e., continuously eliminating deviations).

[0076] Optionally, the pressure sensor adopts a flexible thin film pressure sensor ZNX-01.

[0077] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0078] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A closed-loop control method for a lower limb prosthesis, characterized in that: include: The intrinsic myoelectric signals of the lower limb muscles of the residual limb and the plantar pressure signals of the good leg are collected respectively; identifying a current walking state according to the internal electromyographic signal; According to the current walking state, a control instruction is issued to the prosthesis to drive the prosthesis to move to a specified position; the plantar pressure signal of the good leg is input into the gait recognition model to identify and output the current gait information; According to the current gait information, query the gait plantar pressure information database to obtain the pressure distribution prediction value corresponding to the prosthetic foot; the gait plantar pressure information database includes: plantar pressure distribution data and gait information of the left and right legs during normal walking of the human body; Each current gait information corresponds to multiple plantar pressure signals in the gait plantar pressure information database. The average value of these plantar pressure signals is the pressure distribution prediction value; The pressure signal of the plantar of the prosthetic lower limb is collected, and the pressure signal of the plantar of the prosthetic lower limb is compared with the predicted value of the pressure distribution corresponding to the plantar of the prosthetic limb to obtain a deviation signal; and the prosthetic movement is feedback controlled according to the deviation signal.

2. A closed-loop control method for lower limb prosthesis according to claim 1, characterized in that: The method of identifying the current walking state according to the internal electromyographic signal comprises the following steps: S11, filtering and amplifying the internal electromyographic signal; S12, converting the filtered and amplified internal electromyographic signal into a digital signal through an AD chip; S13, analyzing the digital signal using a fixed sample entropy peak-valley threshold algorithm, extracting peaks and troughs of the internal electromyographic signal sample entropy, and obtaining gait phase information; S14: querying a large gait information database based on the gait phase information to identify the current walking state.

3. A closed-loop control method for lower limb prosthesis according to claim 1, characterized in that: The establishment of the gait recognition model includes the following steps: S21, selecting plantar pressure distribution data and gait information of the left and right legs of a human body during normal walking as input and output data of a pre-built neural network model; the plantar pressure distribution data and gait information of the left and right legs are obtained from the gait plantar pressure information database; S22, iteratively training the neural network model to find the optimal parameters based on the cost equation error of the model; S23. When the cost equation error is minimized, optimal parameters are obtained to obtain a constructed gait recognition model.

4. A closed-loop control method for lower limb prosthesis according to claim 1, characterized in that: The feedback control of the prosthetic limb movement according to the deviation signal includes: processing the deviation signal through a PID algorithm, outputting a corresponding control signal, and performing feedback control on the prosthetic limb movement.

5. A closed-loop control system for a lower limb prosthesis, characterized in that: A closed-loop control method for a lower limb prosthesis according to any one of claims 1 to 4, comprising: a host computer, an electrode wire, a controller, and a pressure sensor; The electrode wire is used to collect the internal electromyographic signals of the residual limb lower limb muscles; The pressure sensor is used to collect the pressure signal of the good leg plantar and the pressure signal of the prosthetic lower limb plantar, and transmit them to the controller; The controller is connected to the host computer, the electrode wire, the pressure sensor and the prosthesis respectively, and is used to wirelessly transmit the internal myoelectric signal to the host computer and issue control instructions to the prosthesis according to the current walking state identified by the host computer; The host computer is used to identify and output the current walking state according to the internal electromyographic signal.

6. The closed-loop control system for lower limb prosthesis according to claim 5, characterized in that: The controller is further configured to wirelessly transmit the plantar pressure signal of the good leg to the host computer, and compare the predicted pressure distribution corresponding to the prosthetic foot identified by the host computer with the plantar pressure signal of the prosthetic lower limb to generate a deviation signal for feedback control of the prosthetic movement; The host computer is further configured to identify and output a predicted pressure distribution value corresponding to the sole of the prosthetic limb based on the sole pressure signal of the good leg.

7. The closed-loop control system for lower limb prosthesis according to claim 5, characterized in that: The pressure sensor adopts the flexible thin film pressure sensor ZNX-01.

Images