Limb prothesis control method and device and storage medium

[0026] First embodiment

[0027] Traditional lower limb prostheses are mainly passive prostheses, which rely on the swing of the residual limb to drive the prosthesis during walking to achieve alternate walking, and the user consumes a lot of energy. The purely active prostheses generally use motors as the executive components to drive joint flexion and extension motions. Because the motors need to run continuously during joint motions, and require a large driving force when standing up, up and down stairs, etc., the power consumption is extremely high. The problem of power supply has become a bottleneck that is difficult to break through. Therefore, the active and passive hybrid drive lower limb prosthesis has become a new research direction at home and abroad. How to set the active control mode and the passive control mode, and how to judge when to use which control mode is a key issue that needs to be solved, which involves sports Type identification problem. To solve this problem, the first embodiment of the present invention provides a prosthetic limb control method, please refer to figure 1 , figure 1 This is a flowchart of a method for controlling a prosthesis according to the first embodiment of the present invention. The method specifically includes the following steps:

[0028] Step S100: Obtain a valid motion signal of the motion performed by the user's limbs.

[0029] To obtain the effective action signal of the action performed by the user's limbs, first obtain the current action signal of the action performed by the user's limbs, and then use a moving window to scan the current action signal and extract the first action signal of the current action signal. A feature value, judging whether the first feature value in the first moving window exceeds a preset threshold, and if yes, judging whether the first feature value in the second moving window exceeds the threshold, When the first characteristic value in the second moving window exceeds the threshold, it is determined that the current action signal within the time from the first moving window to the second moving window is the effective action signal.

[0030] Wherein, the current action signal is a signal that can characterize the user's current action, so the current action signal can be a moving image signal of the user's limbs, posture signals, electromyographic signals, etc., which can distinguish the user's current action Different signals. Most of the current lower limb prostheses rely on sensors such as angle, displacement, force to obtain the motion image signal and posture signal of the user's body movement to identify the specific movement of the user during exercise, so as to realize the recognition of the movement type, but the recognition often lags behind Movement occurs, and the earlier the type of movement is judged, the more accurate the control will be, and the more natural the user will walk with the prosthesis. Electromyography signal (EMG) is the superposition of motor unit action potentials (MUAP) in many muscle fibers in time and space. Surface electromyography signal (SEMG) is the combined effect of superficial muscle EMG and nerve trunk electrical activity on the skin surface. It can reflect neuromuscular activity to a certain extent, and has the advantages of non-invasiveness, non-invasiveness, and simple operation in measurement. The EMG signal has a strong correlation with movement and occurs before the action. Therefore, the use of the EMG signal of the prosthesis wearer to identify the movement intention has become an important way to control the prosthesis. Therefore, as an implementation manner, taking the residual limb of the leg as an example, the effective motion signal and the current motion signal in the first embodiment of the present invention are the surface electromyographic signals of the thigh, including at least the rectus femoris, semitendinosus and gluteus maximus. The electromyographic signals of the three channels of the muscle, optionally, in other embodiments of the present invention may also include the muscles of the superficial muscles of other residual limbs such as the medial femoris, lateral femoris, biceps femoris, and tensor fascia lata. Electric signal channel.

[0031] The EMG signal collected by the existing sensor inevitably contains certain other noise signals. In order to solve this problem, as an implementation manner, a moving window is used to scan the current motion signal and extract the current motion signal. Before the first eigenvalue of, it is also necessary to preprocess the collected EMG signal, that is, perform low-pass filtering on the EMG signal. Further, the low-pass filtering step may be 10-500 Hz Butterworth filtering.

[0032] The moving window moves periodically by the length of time. For example, in this embodiment, 100ms is used as the length of the window. The collection range of the moving window starting at a certain time is the method that will scan the EMG signal within 100ms after the starting time. From now on is the next new moving window. As an implementation manner, when a moving window is used to perform data scanning on the current action signal and extract the first characteristic value of the current action signal, the first characteristic value in this embodiment is the wavelength of the electromyographic signal to move The wavelength of the data in the window is used as a characteristic quantity for judging whether there is an action. The wavelength is defined as Since the wavelength reflects the combined effect of the signal's amplitude, frequency and duration, and can reflect the complexity of the waveform of a segment of myoelectric signal, it is chosen as the basis for judgment. By comparing the wavelength value of the signal in the moving window with the preset threshold, it is judged whether the action has occurred. When the characteristic value exceeds the preset threshold, the moving window is marked as active, otherwise it is marked as inactive. Set the moving window marking state to A(k), then the previous moving window is A(k-1) , The next moving window is A(k+1), and so on, in the moving process of the moving window, when A(k-1) is inactive, A(k) is active, A(k+1) ) Is the active state, A(k+2) is the active state, and when A(k+3) is the inactive state, it is determined that the starting point of the moving window k is the starting point of the group of EMG actions and the moving window (k+3) The end point of) is the end point of the group of EMG actions.

[0033] As another embodiment, after determining that the starting point of the moving window k is the starting point of the group of electromyographic actions, a certain moment after the starting point can be directly set as the ending point of the group of electromyographic actions, among which, it is more reasonable The time can be a certain value between 100-500ms, and the specific settings can be adjusted according to the actual situation.

[0034] At this time, the current action signal between the start point and the end point is determined as the effective action signal.

[0035] Step S200: Identify the current movement type of the action based on the effective action signal.

[0036] Recognizing the current movement type of the action based on the effective action signal in this step specifically includes: extracting a second feature value of the effective action signal; and recognizing the action based on the second feature value and the corresponding relationship The current exercise type. Among them, the root mean square of the EMG signal is considered to be the most reliable parameter in the time domain, and is used to estimate the magnitude of the generated force. The average frequency of the power spectrum of the EMG signal can reflect the energy of the EMG signal. Therefore, the second characteristic value includes at least the root mean square of the EMG signal and the average frequency of the power spectrum at the same time. Optionally, in addition to the above-mentioned root mean square and power spectrum average frequency, when higher accuracy is required or there are other specific requirements, the second characteristic value may also include integrated EMG, standard deviation, wavelength and other time domains. One or more values ​​in frequency domain characteristics such as the characteristic and the median frequency.

[0037] Before identifying the current movement type of the movement based on the effective movement signal, the prosthetic limb control method provided in this embodiment further includes: acquiring the corresponding history of multiple users or multiple testers performing historical exercises of different movement types The historical second characteristic value of the effective action signal, and the corresponding relationship between the movement type and the second characteristic value is established based on the historical second characteristic value. The establishment of the corresponding relationship can be completed by using a machine learning pattern classification algorithm, and a certain amount of sample features of multiple users or multiple testers during historical exercises of different types of sports are extracted in advance for classification training and learning, and the training results are stored. , Perform exercise type recognition based on real-time extracted features and pre-stored training results. Wherein, in the corresponding relationship between the movement type and the second feature value established based on the historical second feature value, the lower limbs are taken as an example. Common types of movement include natural walking, stairs/steps, stairs/steps, sitting For common lower limb movements such as going down and standing up, the user's movements are classified into corresponding movement types according to the second characteristic values ​​extracted from the effective movement signals collected by analysis, so that the movements can be clearly and quickly classified.

[0038] As an implementation manner, the above-mentioned recognition of the current movement type of the action can adopt an online recognition method, using a support vector machine (SVM) as a pattern classifier, or other machines such as K-nearest neighbors, neural networks, hidden Markov models, etc. Learning pattern classification algorithm.

[0039] Step S300: controlling the prosthesis connected with the limb to be in a current matching control mode that matches the current movement type.

[0040] It should be understood that, before performing this step, the prosthetic control method further includes: establishing a matching relationship between the exercise type and the matching control mode based on the energy consumption of the exercise type. Among them, the lower limb movements are taken as an example. Movement types such as natural walking and sitting are low-energy exercises corresponding to the passive control mode, and movement types such as stairs/steps, downstairs/steps, and standing up are high-energy exercises corresponding to the active control mode.

[0041] The prosthetic limb control method provided in this embodiment uses the electromyographic signal to quickly and accurately complete the recognition of the movement type. At the same time, when walking on flat ground, sitting down and other low energy consumption actions, the passive control mode is adopted, and the residual limb is used to drive the prosthetic limb to move, saving the prosthetic limb. Power consumption. When going up and down stairs and standing up and other high energy consumption actions, the active control mode is adopted, and the prosthesis provides power to assist the patient's movement and reduce the patient's energy consumption. Using this combination of active and passive control strategies, compared with traditional passive prostheses, it can reduce the energy consumption of patients wearing prostheses for a long time, and at the same time can solve the problem of insufficient power supply for active prostheses for a long time, which is a smart prosthetic control Provides a new way of thinking.

[0042] Second embodiment

[0043] In order to implement the foregoing information processing method, a second embodiment of the present invention provides a prosthetic limb control device 100. Please refer to figure 2 , figure 2 It is a schematic diagram of a module of a prosthetic limb control device provided by the second embodiment of the present invention. The prosthesis control device 100 includes a signal acquisition module 110, a type recognition module 120 and a mode switching module 130.

[0044] The signal acquisition module 110 is used to acquire the effective motion signal of the movement performed by the user's limbs, and is also used to obtain the current motion signal of the movement performed by the user's limbs.

[0045] The type identification module 120 is configured to identify the current movement type of the action based on the effective action signal.

[0046] The mode switching module 130 is configured to control the prosthesis connected with the limb to be in a current matching control mode that matches the current movement type.

[0047] As an implementation manner, the signal acquisition module 110 provided in this embodiment includes a scanning unit, a first judgment unit, and a second judgment unit.

[0048] The scanning unit is configured to use a moving window to perform data scanning on the current action signal and extract the first characteristic value of the current action signal.

[0049] The first determining unit is configured to determine whether the first characteristic value in the first moving window exceeds a preset threshold.

[0050] The second determining unit is configured to determine whether the first characteristic value in the second moving window exceeds the threshold.

[0051] Further, the type recognition module 130 provided in this embodiment includes a feature extraction unit and a type confirmation unit.

[0052] The feature extraction unit is used to extract the second feature value of the effective action signal.

[0053] The type confirmation unit is configured to identify the current movement type of the action based on the second characteristic value and the corresponding relationship.

[0054] Those skilled in the art can clearly understand that, for the convenience and conciseness of the description, the specific working process of the device described above can refer to the corresponding process in the foregoing method, which will not be repeated here.

[0055] Please refer to image 3 , image 3 It is a structural block diagram of an electronic device that can be applied to the embodiments of the present application provided by the embodiments of the present application. The electronic device 200 may include a prosthesis control device 100, a memory 201, a storage controller 202, a processor 203, a peripheral interface 204, and an input and output unit 205.

[0056] The components of the memory 201, the storage controller 202, the processor 203, the peripheral interface 204, and the input output unit 205 are directly or indirectly electrically connected to each other to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The prosthesis control device 100 includes at least one software function module that can be stored in the memory 201 in the form of software or firmware or solidified in an operating system (OS) of the prosthesis control device 100. The processor 203 is configured to execute an executable module stored in the memory 101, for example, a software function module or a computer program included in the prosthesis control device 100.

[0057] The memory 201 can be, but is not limited to, random access memory (Random Access Memory, RAM), read only memory (Read Only Memory, ROM), programmable read-only memory (Programmable Read-Only Memory, PROM), and erasable Except for Erasable Programmable Read-Only Memory (EPROM), Electric Erasable Programmable Read-Only Memory (EEPROM), etc. Wherein, the memory 201 is used to store a program, and is also used to store the corresponding relationship between the movement type and the second characteristic value, and the matching relationship between the movement type and the matching control mode, wherein the corresponding relationship and the matching relationship can be converted into a database for processing. Storage is more convenient for operation, and the response speed of the prosthetic control method is faster. The processor 203 executes the program after receiving the execution instruction. The method executed by the server defined by the streaming process disclosed in any of the foregoing embodiments of the present invention may be applied to the processor 203, or may be used by the processor 203. achieve.

[0058] The processor 203 may be an integrated circuit chip with signal processing capability. The aforementioned processor 203 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU for short), a network processor (Network Processor, NP), etc.; it may also be a digital signal processor (DSP) or an application specific integrated circuit. (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The methods, steps, and logical block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor 203 may also be any conventional processor or the like.

[0059] The peripheral interface 204 couples various input/output devices to the processor 203 and the memory 201. In some embodiments, the peripheral interface 204, the processor 203, and the storage controller 202 may be implemented in a single chip. In some other instances, they can be implemented by independent chips.

[0060] The input and output unit 205 is used to provide input data for the user to realize the interaction between the user and the server (or local terminal). The input and output unit 205 may be, but is not limited to, a mouse, a keyboard, and the like.

[0061] Understandable, image 3 The structure shown is for illustration only, and the electronic device 200 may also include image 3 More or fewer components shown in the image 3 Different configurations are shown. image 3 The components shown in can be implemented by hardware, software or a combination thereof.

[0062] When the online recognition method is used to recognize the type of movement, the electronic device 200 will also interact with the server, Figure 4 It is a schematic diagram of a user terminal, namely, an electronic device 200 interacting with a server 300 provided by an embodiment of the present invention. The electronic device 200 communicates with one or more electronic devices 200 via a network 400 for data communication or interaction. The server 300 may be a web server, a database server, or the like. The electronic device 200 may be a terminal such as a microcomputer and a wearable device.

[0063] Those skilled in the art can clearly understand that, for the convenience and conciseness of the description, the specific working process of the device described above can refer to the corresponding process in the foregoing method, which will not be repeated here.

[0064] In summary, the embodiments of the present invention provide a prosthetic control method and device. The method first obtains the effective action signal of the action performed by the user's limbs, and identifies the current movement type of the action based on the effective action signal , And then control the prosthesis connected to the limb to be in an active control mode or a passive control mode that matches the current movement type. In the active control mode, the prosthesis provides power to drive the limb to move. In the passive control mode, the limbs provide power to drive the prosthesis movement, which can improve the pattern recognition speed, reduce the energy consumption of the patient wearing the prosthesis for a long time, and can solve the problem of insufficient power supply for the active prosthesis for a long time.