Grabbing method and device based on motion intention and concentration degree, equipment and medium

Abstract

The invention discloses a grabbing method, device and equipment based on a motion intention and concentration degree and a medium, and the method comprises the steps: obtaining a training sample set of an intention level and a grabbing power level, and training a support vector machine model through the training sample set; performing standardization processing on the first energy ratio feature sequence, the first original feature sequence and the first intrinsic mode feature sequence by adopting a Z-Score algorithm to obtain a second energy ratio feature sequence, a second original feature sequence and a plurality of second intrinsic mode feature sequences; and performing classification decision by using the target support vector machine model to obtain an intention level, simultaneously obtaining a user concentration average value, and determining a target output holding power according to the user concentration average value and the intention level. And the stress state of the grabbed target is detected in real time, and PID operation is carried out. Fine hierarchical control of the grabbing force is achieved, the recognition precision of the grabbing intention is improved, and the problem that traditional brain control equipment grabs but cannot stably hold is solved by conducting anti-falling PID operation in real time.

Description

Technical Field

[0001] This invention relates to the field of grasping entertainment equipment technology, and in particular to a grasping method, device, equipment and medium based on motion intention and concentration. Background Technology

[0002] Traditional claw machines and other grasping-type amusement devices primarily rely on joysticks or buttons for physical control, resulting in a limited form of interaction and a lack of immersion. With the development of brain-computer interface technology, simple devices based on concentration control have emerged, but existing technologies suffer from the following significant drawbacks:

[0003] Limited control dimensions: Currently, the "on / off" action is triggered only by the threshold of EEG signals, which cannot achieve fine-grained control of grasping force (such as light grip or tight grip).

[0004] Feature extraction is unstable: the prefrontal electroencephalogram (EEG) signal has significant non-stationarity and is easily interfered with by electrooculogram (EOG) and electromyography (EMG). Traditional time-frequency analysis methods are difficult to extract stable intent features in a short time.

[0005] Lack of a closed-loop force feedback system (fall problem): Most existing brain-controlled devices use open-loop control, meaning that after the brain issues a command, the mechanical gripper outputs a fixed voltage. When an object slips due to gravity or inertia, the system cannot sense this and adjust the gripping force in real time, resulting in a low success rate and a poor user experience. Summary of the Invention

[0006] The purpose of this application is to propose a grasping method, device, equipment, and medium based on motion intention and focus, so as to solve the problems of existing claw machine grasping control methods, which cannot finely control the grasping force, have low intention recognition accuracy, lack anti-falling mechanism, and have poor user experience.

[0007] Firstly, to address the aforementioned issues, a grasping method based on motion intention and focus is proposed, including:

[0008] Step 100: Construct a support vector machine model and set multiple acquisition electrodes on the user's brain. Collect the user's EEG signals at different grasping intention levels according to the specified sampling rate and determine the corresponding grasping force level. Obtain a training sample set of the intention level and the grasping force level. Use the training sample set to train the support vector machine model to obtain the target support vector machine model.

[0009] Step 200: Acquire real-time EEG signals from the corresponding channel electrodes using the acquisition electrodes. Preprocess the real-time EEG signals to obtain preprocessed EEG signals. Decompose the preprocessed EEG signals using the empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components. Calculate the first energy ratio feature sequence of the current channel using the multiple intrinsic mode signal components. Extract features from the preprocessed EEG signals and the multiple intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence. Standardize the first energy ratio feature sequence, the first original feature sequence, and the first intrinsic mode feature sequence using the Z-Score algorithm to obtain a second energy ratio feature sequence, a second original feature sequence, and multiple second intrinsic mode feature sequences. The preprocessed EEG signals include... frequency band frequency band and Frequency band;

[0010] Step 300: Construct an input feature vector using the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic modality feature sequences. Input this vector into the target support vector machine model for classification decision-making to obtain the intent level. Simultaneously, acquire the preprocessed EEG signal. Frequency band signal power, according to the The average user attention level is obtained by measuring the frequency band signal power, and the target output grip force is determined based on the average user attention level and the intent level.

[0011] Step 400: Real-time detection of the force state of the grasped target and PID calculation to obtain the difference between the actual output grasping force and the target output grasping force in real time. If the grasped target is detected to be sliding or the grasping force difference is greater than a specified threshold, the output grasping force is adjusted to prevent the target from falling.

[0012] Furthermore, step 100 includes:

[0013] Step 110: Obtain multiple grip strength levels and determine the intent level corresponding to each grip strength level;

[0014] Step 120: Collect the EEG signal corresponding to the user's intention level, extract features, and construct a training input vector using the extracted feature sequence;

[0015] Step 130: Input the training input vector and the corresponding gripping force level information into the support vector machine model for training to obtain the target support vector machine model.

[0016] Furthermore, step 200 includes:

[0017] Step 210: Filter the EEG signal using a Butterworth bandpass filter to retain the relevant information from the EEG signal. frequency band frequency band and Frequency band;

[0018] Step 220: Obtain the signal voltage of the EEG signal. If the signal voltage is greater than a specified threshold, perform interpolation smoothing on the EEG signal to remove artifacts.

[0019] Furthermore, after step 220, step 200 also includes:

[0020] Step 230: Calculate the envelope mean of the preprocessed EEG signal, and use the envelope mean to calculate the intermediate signal of the preprocessed EEG signal;

[0021] Step 240: Calculate the first intrinsic mode signal component based on the intermediate signal;

[0022] Step 250: Calculate the first residual signal based on the first intrinsic modal signal component and the preprocessed EEG signal, and use the first residual signal as the input signal to calculate the next intrinsic modal signal component.

[0023] Furthermore, after step 250, step 200 also includes:

[0024] Step 260: Calculate the component energy of each intrinsic mode signal component;

[0025] Step 270: Calculate the energy ratio using a specified number of adjacent component energies, and calculate the first energy characteristic ratio sequence using the energy ratios of all intrinsic mode signal components.

[0026] Furthermore, step 300 includes:

[0027] Step 310: Obtain the first gripping force level corresponding to the intention level, and obtain the average user attention level of the current EEG signal;

[0028] Step 320: Determine the focus energy of the current intent level based on the average user focus level. If the focus energy does not reach the specified threshold, then use the first gripping force level as the target output gripping force.

[0029] Step 330: If the focused energy reaches the specified threshold, the maximum gripping force level is used as the target output gripping force.

[0030] Furthermore, step 400 includes:

[0031] Step 410: Update the output gripping force in real time at a first predetermined frequency, and perform PID calculation at a second predetermined frequency to obtain the gripping force difference;

[0032] Step 420: Determine whether the gripping force difference exceeds the specified threshold. If it does, adjust the output voltage until the gripping force difference is less than the specified threshold.

[0033] Wherein, the first specified frequency is less than the second specified frequency.

[0034] To address the aforementioned problems, a grasping device based on motion intent and focus is provided, employing the aforementioned grasping method based on motion intent and focus, comprising:

[0035] The training module is used to construct a support vector machine model and set multiple acquisition electrodes in the user's brain. It collects the user's EEG signals at different grasping intention levels according to a specified sampling rate and determines the corresponding grasping force level. It obtains a training sample set of the intention level and the grasping force level, and uses the training sample set to train the support vector machine model to obtain the target support vector machine model.

[0036] The feature extraction module is used to acquire real-time EEG signals from corresponding channel electrodes using the acquisition electrodes, preprocess the real-time EEG signals to obtain preprocessed EEG signals, decompose the preprocessed EEG signals using an empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components, calculate the first energy ratio feature sequence of the current channel using the multiple intrinsic mode signal components, extract features from the preprocessed EEG signals and the multiple intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence, and perform Z-Score algorithm to standardize the first energy ratio feature sequence, the first original feature sequence, and the first intrinsic mode feature sequence to obtain a second energy ratio feature sequence, a second original feature sequence, and multiple second intrinsic mode feature sequences. The preprocessed EEG signals include... frequency band frequency band and Frequency band;

[0037] The output module is used to construct an input feature vector using the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic mode feature sequences, input the vector into the target support vector machine model for classification decision-making, obtain the intent level, and simultaneously acquire the preprocessed EEG signal. Frequency band signal power, according to the The average user attention level is obtained by measuring the frequency band signal power, and the target output grip force is determined based on the average user attention level and the intent level.

[0038] The PID module is used to detect the force state of the grasped target in real time and perform PID calculations to obtain the difference between the actual output grasping force and the target output grasping force in real time. If the grasped target is detected to be sliding or the grasping force difference is greater than a specified threshold, the output grasping force is adjusted to prevent the target from falling.

[0039] To address the aforementioned technical problems, this application also provides an electronic device, including a memory and a processor. The memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the grasping method based on motion intent and focus as described in the first aspect.

[0040] To address the aforementioned technical problems, embodiments of this application also provide a computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps of the grasping method based on motion intent and focus as described in the first aspect.

[0041] Compared with the prior art, the embodiments of this application have the following main technical effects: by acquiring a training sample set of intent level and grasping force level, the support vector machine model is trained, the preprocessed EEG signal is decomposed using the empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components, and an input feature vector is constructed based on the second energy ratio feature sequence, the second original feature sequence and multiple second intrinsic mode feature sequences, thereby realizing fine-level control of grasping force, improving the recognition accuracy of grasping intent, and improving the rapid adjustment capability by performing anti-drop PID calculation in real time, thus solving the problem of traditional brain control devices "can grasp but cannot hold firmly". Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram illustrating a specific implementation of a grasping method based on motion intent and focus in this invention.

[0044] Figure 2 for Figure 1 A schematic diagram of a specific implementation of step 100 in the diagram;

[0045] Figure 3 for Figure 1 A schematic diagram of a specific implementation of step 200 in the diagram;

[0046] Figure 4 for Figure 3 A schematic diagram of a specific implementation method following step 220;

[0047] Figure 5 for Figure 4 A schematic diagram of a specific implementation method following step 250;

[0048] Figure 6 for Figure 5 A schematic diagram of a specific implementation of step 300 in the process;

[0049] Figure 7 for Figure 6 A schematic diagram of a specific implementation of step 400 in the diagram;

[0050] Figure 8 This is a schematic diagram of the structure of a grasping device based on motion intention and focus in this invention;

[0051] Figure 9 This is a schematic diagram of the structure of an embodiment of an electronic device according to this application. Detailed Implementation

[0052] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are all within the scope of protection of this invention.

[0053] In this document, the term "embodiment" means that a particular 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 separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings of this application are intended to cover non-exclusive inclusion.

[0055] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0056] The purpose of this application is to propose a grasping method, device, equipment, and medium based on motion intention and focus, so as to solve the problems of existing claw machine grasping control methods, which cannot finely control the grasping force, have low intention recognition accuracy, lack anti-falling mechanism, and have poor user experience.

[0057] Firstly, to address the aforementioned issues, a grasping method based on motion intention and focus is proposed, such as... Figure 1 , Figure 1 This is a schematic diagram illustrating a specific implementation of a grasping method based on motion intent and focus according to the present invention; including:

[0058] Step 100: Construct a support vector machine model and set up multiple acquisition electrodes on the user's brain. Collect EEG signals of the user at different grasping intention levels according to the specified sampling rate and determine the corresponding grasping force level. Obtain a training sample set of intention level and grasping force level. Use the training sample set to train the support vector machine model to obtain the target support vector machine model.

[0059] In one possible implementation, such as Figure 2 , Figure 2 for Figure 1 A schematic diagram of a specific implementation of step 100; step 100 includes:

[0060] Step 110: Obtain multiple grip strength levels and determine the corresponding intent level; Step 120: Collect EEG signals corresponding to the user's intent level, extract features, and construct a training input vector using the extracted feature sequences; Step 130: Input the training input vector and the corresponding grip strength level information into the support vector machine model for training to obtain the target support vector machine model.

[0061] In this embodiment, a dual-channel EEG device can be used, with the channel electrodes placed at the Fp1 and Fp2 positions of the prefrontal cortex (international 10-20 electrode system standard). This device can stably acquire precise EEG signals from brain regions related to human cognition and thought functions.

[0062] In this embodiment, the kernel function of the Support Vector Machine (SVM) model adopts the Radial Basis Function (RBF) to avoid linear inseparability of features in low-dimensional space. Five grip strength levels can be set (very light 1, light grip 2, medium grip 3, slightly tight grip 4, tight grip 5), corresponding to five intention levels. During operation, the operator needs to maintain the corresponding grip intention (intention level). For example, grip strength level 1 corresponds to the operator imagining a gentle control of an object, and level 4 corresponds to the operator imagining a forceful control of an object. Twenty sets of data are collected for each grip strength level (each set of data lasts 3 seconds, with a sampling rate of 256Hz) to form a training sample set for the intention level and grip strength level.

[0063] Step 200: Acquire real-time EEG signals from the corresponding channel electrodes using the acquisition electrodes. Preprocess the real-time EEG signals to obtain preprocessed EEG signals. Decompose the preprocessed EEG signals using the Empirical Mode Decomposition (EMD) algorithm to obtain multiple intrinsic mode signal components. Calculate the first energy ratio feature sequence of the current channel using the multiple intrinsic mode signal components. Extract features from the preprocessed EEG signals and the multiple intrinsic mode signal components to obtain the first original feature sequence and the first intrinsic mode feature sequence. Standardize the first energy ratio feature sequence, the first original feature sequence, and the first intrinsic mode feature sequence using the Z-Score algorithm to obtain the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic mode feature sequences. The preprocessed EEG signals include... frequency band frequency band and Frequency band.

[0064] In one possible implementation, such as Figure 3 , Figure 3 for Figure 1 A schematic diagram of a specific implementation of step 200; step 200 includes:

[0065] Step 210: Filter the EEG signal using a Butterworth bandpass filter to retain the relevant information from the EEG signal. frequency band frequency band and Frequency band; Step 220: Obtain the signal voltage of the EEG signal. If the signal voltage is greater than the specified threshold, perform interpolation smoothing on the EEG signal to remove artifacts.

[0066] In this embodiment, the acquired EEG signals are passed through a Butterworth bandpass filter of 3-45Hz to retain the frequency bands of alpha, beta, and gamma waves, while filtering out 50Hz power frequency interference and low-frequency baseline drift, and removing artifacts.

[0067] In one possible implementation, after step 220, as Figure 4 , Figure 4 for Figure 3 A schematic diagram of a specific implementation following step 220; step 200 further includes:

[0068] Step 230: Calculate the envelope mean of the preprocessed EEG signal, and use the envelope mean to calculate the intermediate signal of the preprocessed EEG signal; Step 240: Calculate the first intrinsic mode signal component based on the intermediate signal; Step 250: Calculate the first residual signal based on the first intrinsic mode signal component and the preprocessed EEG signal, and use the first residual signal as the input signal to calculate the next intrinsic mode signal component.

[0069] In this embodiment, the Empirical Mode Decomposition (EMD) algorithm is used to decompose the EEG signal x(t). The EEG signal is a non-stationary and nonlinear signal. By definition, the Empirical Mode Decomposition (EMD) algorithm needs to satisfy two key conditions:

[0070] Throughout the entire signal range, the number of poles is equal to or differs from the number of zero crossings by at most one.

[0071] At any given time, the mean of the upper envelope (maximum fitting) and lower envelope (minimum fitting) of the signal is 0.

[0072] First, calculate the upper and lower envelopes, then calculate the envelope mean, remove the envelope mean, and obtain the above intermediate signal. Repeat the above steps until the above two key conditions are met. At this time, the first intrinsic mode signal component IMF1=C1(t) is the highest frequency IMF.

[0073] For subsequent IMF extraction, the first residual signal r1(t) = x(t) - C1(t) is calculated first. Using the first residual signal r1(t) as the input signal, the filtering is repeated to obtain the next intrinsic mode signal component IMF2.

[0074] In one possible implementation, after step 250, as Figure 5 , Figure 5 for Figure 4 A schematic diagram of a specific implementation following step 250; step 200 further includes:

[0075] Step 260: Calculate the component energy of each intrinsic mode signal component; Step 270: Calculate the energy ratio using a specified number of adjacent component energies, and calculate the first energy characteristic ratio sequence using the energy ratios of all intrinsic mode signal components.

[0076] In this embodiment, amplitude integration is performed on multiple intrinsic mode signal components to obtain energies E1, E2, ..., En, respectively. Four adjacent component energies are selected, and the energy ratio is:

[0077] E K=E K1 +E K2 / E K3 +E K4 (1)

[0078] The first energy characteristic ratio sequence is obtained by continuously performing the above calculations.

[0079] In this embodiment, component energy and frequency band frequency band and The frequency band relationship is as follows:

[0080] Empirical mode decomposition (EMD) was performed on the prefrontal EEG signals. After signal decomposition, multiple IMFs were obtained. When β and γ waves were dominant, the energy of the high-frequency IMF components (IMF1, IMF2) increased. When α waves were dominant, the energy of the mid-frequency IMF components (IMF3, IMF4) was relatively high.

[0081] Step 300: Construct an input feature vector using the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic modal feature sequences. Input this vector into the target support vector machine model for classification decision-making to obtain the intent level. Simultaneously, acquire the β-band signal power of the preprocessed EEG signal. The average user attention level is obtained by measuring the frequency band signal power, and the target output gripping force is determined based on the average user attention level and the level of intent.

[0082] In one possible implementation, such as Figure 6 , Figure 6 for Figure 5 A schematic diagram of a specific implementation of step 300; step 300 includes:

[0083] Step 310: Obtain the first gripping force level corresponding to the intention level, and obtain the average user attention level of the current EEG signal; Step 320: Determine the attention energy of the current intention level based on the average user attention level. If the attention energy does not reach the specified threshold, the first gripping force level is used as the target output gripping force; Step 330: If the attention energy reaches the specified threshold, the maximum gripping force level is used as the target output gripping force.

[0084] In this embodiment, the gripping force can be controlled by the output voltage. Simultaneously, the operator's average concentration level is calculated every 5 seconds. The stronger the β-wave power, the higher the brain's concentration. When the average concentration level is greater than 5, the concentration energy increases by 1; when it's greater than 40, it increases by 2; when it's greater than 60, it increases by 3; when it's greater than 80, it increases by 4; and when the accumulated concentration energy exceeds 90, the energy bar is fully lit, allowing the operator to grip the doll with maximum force.

[0085] Step 400: Real-time detection of the force state of the target being grasped and PID calculation to obtain the difference between the actual output grasping force and the target output grasping force. If the target is detected to be slipping or the difference in grasping force is greater than the specified threshold, the output grasping force is adjusted to prevent the target from falling.

[0086] In one possible implementation, such as Figure 7 , Figure 7 for Figure 6 A schematic diagram of a specific implementation of step 400; step 400 includes:

[0087] Step 410: Update the output gripping force in real time at a first specified frequency, and perform PID calculation at a second specified frequency to obtain the gripping force difference; Step 420: Determine whether the gripping force difference exceeds a specified threshold. If it does, adjust the output voltage until the gripping force difference is less than the specified threshold.

[0088] The first specified frequency is less than the second specified frequency.

[0089] In this embodiment, the first specified frequency can be 1Hz, and the second specified frequency can be 100Hz. The slow intent + fast execution method is used to prevent the target object from falling.

[0090] In this embodiment, a support vector machine model is trained by acquiring a training sample set of intent level and grasping force level. The preprocessed EEG signal is decomposed using the empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components. An input feature vector is constructed based on the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic mode feature sequences. This achieves fine-grained hierarchical control of grasping force, improves the recognition accuracy of grasping intent, and enhances the rapid adjustment capability by performing anti-drop PID calculations in real time. This solves the problem of traditional brain-controlled devices being able to "grasp but not hold firmly".

[0091] To address the aforementioned problems, a grasping device based on motion intention and focus is provided, employing a grasping method based on motion intention and focus, such as... Figure 8 , Figure 8 This is a schematic diagram of the structure of a grasping device based on motion intention and focus according to the present invention; it includes:

[0092] The training module 501 is used to construct a support vector machine model and set multiple acquisition electrodes in the user's brain. It collects the user's EEG signals at different grasping intention levels according to the specified sampling rate and determines the corresponding grasping force level. It obtains a training sample set of intention level and grasping force level, and uses the training sample set to train the support vector machine model to obtain the target support vector machine model.

[0093] Feature extraction module 502 is used to acquire real-time EEG signals from corresponding channel electrodes using acquisition electrodes, preprocess the real-time EEG signals to obtain preprocessed EEG signals, decompose the preprocessed EEG signals using an empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components, calculate the first energy ratio feature sequence of the current channel using the multiple intrinsic mode signal components, extract features from the preprocessed EEG signals and the multiple intrinsic mode signal components respectively to obtain a first original feature sequence and a first intrinsic mode feature sequence, and standardize the first energy ratio feature sequence, the first original feature sequence and the first intrinsic mode feature sequence using the Z-Score algorithm respectively to obtain a second energy ratio feature sequence, a second original feature sequence and multiple second intrinsic mode feature sequences. The preprocessed EEG signals include frequency band frequency band and Frequency band;

[0094] Output module 503 is used to construct an input feature vector using the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic mode feature sequences. This vector is then input into the target support vector machine model for classification decision-making to obtain the intent level. Simultaneously, it acquires preprocessed EEG signals. Frequency band signal power, according to The average user attention level is obtained by measuring the frequency band signal power, and the target output gripping force is determined based on the average user attention level and the level of intent.

[0095] The PID module 504 is used to detect the force state of the grasped target in real time and perform PID calculations to obtain the difference between the actual output grasping force and the target output grasping force in real time. If the grasped target is detected to be slipping or the grasping force difference is greater than a specified threshold, the output grasping force is adjusted to prevent the target from falling.

[0096] To address the aforementioned technical problems, this application also provides an electronic device, including a memory and a processor. The memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the grasping method based on motion intent and focus as described in the first aspect.

[0097] The following technical solution was adopted: it includes a processor, a network module, and a memory, with the processor and memory interconnected through the network module.

[0098] This electronic device can be a computer, server, workstation, or other similar device; it can also be a mobile device such as a mobile phone, tablet, or in-vehicle mobile terminal; or other devices with program execution capabilities. Its internal structure diagram can be as follows: Figure 9 As shown, Figure 9This is a schematic diagram of an embodiment of an electronic device according to this application. The electronic device includes a processor, a memory, and a network module. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores an operating system, instructions, or code. The internal memory provides an environment for the operation of the operating system and instructions or code in the non-volatile storage media. When the instructions or code are executed by the processor, they implement the functions or steps of the aforementioned grasping method based on motion intent and attention. The network module of the electronic device may include a network interface and / or a wireless network module, allowing the electronic device to communicate with other devices or service platforms. Furthermore, the electronic device may also include a display screen and input devices, etc.

[0099] The memory is used to store computer programs, which include program instructions. The processor is configured to invoke the program instructions, and when the processor executes the instructions or code, it implements the steps of the grasping method based on motion intent and focus as described above.

[0100] To address the aforementioned technical problems, embodiments of this application also provide a computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps of the grasping method based on motion intent and focus as described in the first aspect.

[0101] The aforementioned computer-readable storage medium can be a device for a grasping method based on motion intent and focus provided in any of the foregoing embodiments, or an internal storage unit of the aforementioned terminal device, such as a hard drive or memory of an electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device.

[0102] Furthermore, the computer-readable storage medium may include both internal storage units and external storage devices of the electronic device. The computer-readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0103] However, it should be understood that it is not required to implement all the components shown; more or fewer components may be implemented instead. Those skilled in the art will understand that the electronic device described herein is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0104] The electronic device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The electronic device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0105] Compared with the prior art, the embodiments of this application have the following main technical effects: by acquiring a training sample set of intent level and grasping force level, the support vector machine model is trained, the preprocessed EEG signal is decomposed using the empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components, and an input feature vector is constructed based on the second energy ratio feature sequence, the second original feature sequence and multiple second intrinsic mode feature sequences, thereby realizing fine-level control of grasping force, improving the recognition accuracy of grasping intent, and improving the rapid adjustment capability by performing anti-drop PID calculation in real time, thus solving the problem of traditional brain control devices "can grasp but cannot hold firmly".

[0106] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A grasping method based on motion intention and focus, characterized in that it includes: Step 100: Construct a support vector machine model and set multiple acquisition electrodes on the user's brain. Collect the user's EEG signals at different grasping intention levels according to the specified sampling rate and determine the corresponding grasping force level. Obtain a training sample set of the intention level and the grasping force level. Use the training sample set to train the support vector machine model to obtain the target support vector machine model. Step 200: Acquire real-time EEG signals from the corresponding channel electrodes using the acquisition electrodes. Preprocess the real-time EEG signals to obtain preprocessed EEG signals. Decompose the preprocessed EEG signals using the empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components. Calculate the first energy ratio feature sequence of the current channel using the multiple intrinsic mode signal components. Extract features from the preprocessed EEG signals and the multiple intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence. Standardize the first energy ratio feature sequence, the first original feature sequence, and the first intrinsic mode feature sequence using the Z-Score algorithm to obtain a second energy ratio feature sequence, a second original feature sequence, and multiple second intrinsic mode feature sequences. The preprocessed EEG signals include... frequency band frequency band and Frequency band; Step 300: Construct an input feature vector using the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic modality feature sequences. Input this vector into the target support vector machine model for classification decision-making to obtain the intent level. Simultaneously, acquire the preprocessed EEG signal. Frequency band signal power, according to the The average user attention level is obtained from the frequency band signal power, and the target output grip force is determined based on the average user attention level and the intent level. Step 400: Real-time detection of the force state of the grasped target and PID calculation to obtain the difference between the actual output grasping force and the target output grasping force in real time. If the grasped target is detected to be sliding or the grasping force difference is greater than a specified threshold, the output grasping force is adjusted to prevent the target from falling.

2. The grasping method based on motion intention and focus according to claim 1, characterized in that, Step 100 includes: Step 110: Obtain multiple grip strength levels and determine the intent level corresponding to each grip strength level; Step 120: Collect the EEG signal corresponding to the user's intention level, extract features, and construct a training input vector using the extracted feature sequence; Step 130: Input the training input vector and the corresponding gripping force level information into the support vector machine model for training to obtain the target support vector machine model.

3. The grasping method based on motion intention and focus according to claim 1, characterized in that, Step 200 includes: Step 210: Filter the real-time EEG signal using a Butterworth bandpass filter to retain the relevant information from the real-time EEG signal. frequency band frequency band and Frequency band; Step 220: Obtain the signal voltage of the EEG signal. If the signal voltage is greater than a specified threshold, perform interpolation smoothing on the EEG signal to remove artifacts.

4. The grasping method based on motion intention and focus according to claim 3, characterized in that, Following step 220, step 200 further includes: Step 230: Calculate the envelope mean of the preprocessed EEG signal, and use the envelope mean to calculate the intermediate signal of the preprocessed EEG signal; Step 240: Calculate the first intrinsic mode signal component based on the intermediate signal; Step 250: Calculate the first residual signal based on the first intrinsic modal signal component and the preprocessed EEG signal, and use the first residual signal as the input signal to calculate the next intrinsic modal signal component.

5. The grasping method based on motion intention and focus according to claim 4, characterized in that, After step 250, step 200 further includes: Step 260: Calculate the component energy of each intrinsic mode signal component; Step 270: Calculate the energy ratio using a specified number of adjacent component energies, and calculate the first energy characteristic ratio sequence using the energy ratios of all intrinsic mode signal components.

6. The grasping method based on motion intention and focus according to claim 1, characterized in that, Step 300 includes: Step 310: Obtain the first gripping force level corresponding to the intention level, and obtain the average user attention level of the current EEG signal; Step 320: Determine the focus energy of the current intent level based on the average user focus level. If the focus energy does not reach the specified threshold, then use the first gripping force level as the target output gripping force. Step 330: If the focused energy reaches the specified threshold, the maximum gripping force level is used as the target output gripping force.

7. The grasping method based on motion intention and focus according to claim 1, characterized in that, Step 400 includes: Step 410: Update the output gripping force in real time at a first predetermined frequency, and perform PID calculation at a second predetermined frequency to obtain the gripping force difference; Step 420: Determine whether the gripping force difference exceeds the specified threshold. If it does, adjust the output voltage until the gripping force difference is less than the specified threshold. Wherein, the first specified frequency is less than the second specified frequency.

8. A grasping device based on motion intention and focus, employing the grasping method based on motion intention and focus as described in any one of claims 1-7, characterized in that, include: The training module is used to construct a support vector machine model and set multiple acquisition electrodes in the user's brain. It collects the user's EEG signals at different grasping intention levels according to a specified sampling rate and determines the corresponding grasping force level. It obtains a training sample set of the intention level and the grasping force level, and uses the training sample set to train the support vector machine model to obtain the target support vector machine model. The feature extraction module is used to acquire real-time EEG signals from corresponding channel electrodes using the acquisition electrodes, preprocess the real-time EEG signals to obtain preprocessed EEG signals, decompose the preprocessed EEG signals using an empirical mode decomposition algorithm to obtain multiple intrinsic mode signal components, calculate the first energy ratio feature sequence of the current channel using the multiple intrinsic mode signal components, extract features from the preprocessed EEG signals and the multiple intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence, and perform Z-Score algorithm to standardize the first energy ratio feature sequence, the first original feature sequence, and the first intrinsic mode feature sequence to obtain a second energy ratio feature sequence, a second original feature sequence, and multiple second intrinsic mode feature sequences. The preprocessed EEG signals include... frequency band frequency band and Frequency band; The output module is used to construct an input feature vector using the second energy ratio feature sequence, the second original feature sequence, and multiple second intrinsic mode feature sequences, input the vector into the target support vector machine model for classification decision-making, obtain the intent level, and simultaneously acquire the preprocessed EEG signal. Frequency band signal power, according to the The average user attention level is obtained from the frequency band signal power, and the target output grip force is determined based on the average user attention level and the intent level. The PID module is used to detect the force state of the grasped target in real time and perform PID calculations to obtain the difference between the actual output grasping force and the target output grasping force in real time. If the grasped target is detected to be sliding or the grasping force difference is greater than a specified threshold, the output grasping force is adjusted to prevent the target from falling.

9. An electronic device, characterized in that, The device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the grasping method based on motion intent and focus as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the grasping method based on motion intent and focus as described in any one of claims 1 to 7.

Images