Search

CN-122018686-A - Grabbing method, device, equipment and medium based on exercise intention and concentration degree

CN122018686ACN 122018686 ACN122018686 ACN 122018686ACN-122018686-A

Abstract

The application discloses a grabbing method, a device, equipment and a medium based on movement intention and concentration, wherein the method comprises the steps of obtaining a training sample set of intention level and grabbing level, training a support vector machine model by using the training sample set, respectively carrying out standardization processing on a first energy bit feature sequence, a first original feature sequence and a first intrinsic mode feature sequence by adopting a Z-Score algorithm to obtain a second energy bit feature sequence, a second original feature sequence and a plurality of second intrinsic mode feature sequences, carrying out classification decision by using a target support vector machine model to obtain the intention level, simultaneously obtaining a user concentration average value, and determining target output grabbing force according to the user concentration average value and the intention level. And detecting the stress state of the grabbing target in real time and performing PID operation. The precise hierarchical control of the grabbing force is realized, the identification precision of the grabbing intention is improved, and the problem that the traditional brain control equipment is not stable in grabbing is solved by carrying out anti-drop PID operation in real time.

Inventors

  • HUANG SHAOJIA
  • Xiang Weiwen
  • MAO JIDONG
  • LIN YUCHEN

Assignees

  • 深圳水母智脑科技有限公司

Dates

Publication Date
20260512
Application Date
20260126

Claims (10)

  1. 1. A method of grasping based on movement intent and concentration, comprising: Step 100, constructing a support vector machine model, setting a plurality of acquisition electrodes in the brain of a user, acquiring electroencephalogram signals of the user at different grabbing intention levels according to a specified sampling rate, determining corresponding grabbing force levels, acquiring a training sample set of the intention levels and the grabbing force levels, and training the support vector machine model by using the training sample set to obtain a target support vector machine model; Step 200, collecting a real-time electroencephalogram signal of a corresponding channel electrode by using the collecting electrode, preprocessing the real-time electroencephalogram signal to obtain a preprocessed electroencephalogram signal, decomposing the preprocessed electroencephalogram signal by using an empirical mode decomposition algorithm to obtain a plurality of intrinsic mode signal components, calculating by using the plurality of intrinsic mode signal components to obtain a first energy sign sequence of a current channel, respectively carrying out feature extraction on the preprocessed electroencephalogram signal and the plurality of intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence, respectively carrying out standardization processing on the first energy sign sequence, the first original feature sequence and the first intrinsic mode feature sequence by using a Z-Score algorithm to obtain a second energy sign sequence, a second original feature sequence and a plurality of second intrinsic mode feature sequences, wherein the preprocessed electroencephalogram signal comprises Frequency band(s), Frequency band Frequency band; Step 300, constructing an input feature vector by using the second energy bit sequence, the second original feature sequence and a plurality of second eigenvalue feature sequences, inputting the target support vector machine model for classification decision to obtain an intention level, and simultaneously obtaining the preprocessed electroencephalogram signals Frequency band signal power according to the Acquiring a user concentration degree average value by frequency band signal power, and determining target output grabbing force according to the user concentration degree average value and the intention level; Step 400, detecting the stress state of the grabbing target in real time, performing PID operation to obtain the grabbing force difference between the actual output grabbing force and the target output grabbing force in real time, and if the grabbing target is detected to slide or the grabbing force difference is larger than a specified threshold value, adjusting the output grabbing force to prevent the target from falling.
  2. 2. The exercise intent and concentration based gripping method according to claim 1, wherein said step 100 includes: step 110, acquiring a plurality of gripping force levels, and determining an intention level corresponding to the gripping force level; step 120, acquiring an electroencephalogram signal corresponding to the intention level of the user, extracting features, and constructing a training input vector by using the extracted feature sequence; And 130, inputting the training input vector and the corresponding gripping power level information into the support vector machine model for training to obtain the target support vector machine model.
  3. 3. The exercise intent and concentration based gripping method according to claim 1, wherein said step 200 includes: step 210, filtering the real-time electroencephalogram signal by using a Butterworth band-pass filter, and reserving the real-time electroencephalogram signal Frequency band(s), Frequency band Frequency band; step 220, obtaining the signal voltage of the electroencephalogram signal, and if the signal voltage is greater than a specified threshold, performing interpolation smoothing processing on the electroencephalogram signal to remove artifacts.
  4. 4. The exercise intent and concentration based gripping method according to claim 3, wherein after said step 220, said step 200 further includes: Step 230, calculating an envelope mean value of the preprocessed electroencephalogram signals, and calculating intermediate signals of the preprocessed electroencephalogram signals by using the envelope mean value; Step 240, calculating to obtain a first intrinsic mode signal component according to the intermediate signal; Step 250, calculating and obtaining a first residual signal according to the first intrinsic mode signal component and the preprocessed electroencephalogram signal, and calculating a next intrinsic mode signal component by taking the first residual signal as an input signal.
  5. 5. The exercise intent and concentration based gripping method according to claim 4, wherein after said step 250, said step 200 further includes: Step 260, calculating the component energy of each of the eigenmode signal components; Step 270, calculating an energy ratio by using a specified number of adjacent component energies, and calculating the first energy feature ratio sequence by using the energy ratios of all the eigenvalue signal components.
  6. 6. The exercise intent and concentration based gripping method according to claim 1, wherein said step 300 includes: step 310, obtaining a first grasping power level corresponding to the intention level, and obtaining a user concentration average value of the current electroencephalogram signal; Step 320, determining concentration energy of the current intention level according to the user concentration average value, and if the concentration energy does not reach a specified threshold, taking the first gripping power level as a target output gripping power; and 330, if the concentration energy reaches a specified threshold, taking the maximum gripping power level as a target output gripping power.
  7. 7. The exercise intent and concentration based gripping method according to claim 1, wherein the step 400 includes: Step 410, updating the output gripping power in real time at a first specified frequency, and performing PID operation at a second specified frequency to obtain a gripping power difference; Step 420, judging whether the gripping power difference exceeds a specified threshold, and if so, adjusting the output voltage until the gripping power difference is smaller than the specified threshold; wherein the first prescribed frequency is smaller than the second prescribed frequency.
  8. 8. A gripping apparatus based on exercise intention and concentration, employing a gripping method based on exercise intention and concentration as claimed in any one of claims 1 to 7, characterized by comprising: The training module is used for constructing a support vector machine model, arranging a plurality of acquisition electrodes on the brain of a user, acquiring electroencephalogram signals of the user at different grabbing intention levels according to a specified sampling rate, determining corresponding grabbing force levels, acquiring a training sample set of the intention levels and the grabbing force levels, and training the support vector machine model by using the training sample set to obtain a target support vector machine model; The feature extraction module is used for acquiring real-time electroencephalogram signals of the corresponding channel electrodes by using the acquisition electrodes, preprocessing the real-time electroencephalogram signals to obtain preprocessed electroencephalogram signals, decomposing the preprocessed electroencephalogram signals by using an empirical mode decomposition algorithm to obtain a plurality of intrinsic mode signal components, calculating by using the plurality of intrinsic mode signal components to obtain a first energy bit feature sequence of the current channel, respectively carrying out feature extraction on the preprocessed electroencephalogram signals and the plurality of intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence, respectively carrying out standardization processing on the first energy bit sequence, the first original feature sequence and the first intrinsic mode feature sequence by using a Z-Score algorithm to obtain a second energy bit feature sequence, a second original feature sequence and a plurality of second intrinsic mode feature sequences, wherein the preprocessed electroencephalogram signals comprise Frequency band(s), Frequency band Frequency band; the output module is used for constructing an input feature vector by using the second energy bit sequence, the second original feature sequence and the plurality of second eigenvalue feature sequences, inputting the target support vector machine model for classification decision to obtain the intention level, and simultaneously obtaining the preprocessed electroencephalogram signals Frequency band signal power according to the Acquiring a user concentration degree average value by frequency band signal power, and determining target output grabbing force according to the user concentration degree average value and the intention level; And the PID module is used for detecting the stress state of the grabbing target in real time and carrying out PID operation to obtain the grabbing force difference between the actual output grabbing force and the target output grabbing force in real time, and if the grabbing target is detected to slide or the grabbing force difference is larger than a specified threshold value, the output grabbing force is regulated so as to prevent the target from falling.
  9. 9. An electronic device comprising a memory and a processor, the memory having stored therein computer readable instructions which when executed by the processor implement the steps of the exercise intent and concentration based crawling method of any of claims 1 to 7.
  10. 10. A computer readable storage medium, characterized in that it has stored thereon computer readable instructions which, when executed by a processor, implement the steps of the exercise intention and concentration based gripping method according to any of claims 1 to 7.

Description

Grabbing method, device, equipment and medium based on exercise intention and concentration degree Technical Field The invention relates to the technical field of grabbing entertainment equipment, in particular to a grabbing method, a grabbing device, grabbing equipment and grabbing media based on sports intention and concentration. Background Traditional baby machines or grabbing entertainment equipment mainly rely on rockers or buttons to perform physical control, and are single in interaction form and lack of immersion. With the development of brain-computer interface technology, simple devices based on concentration (Concentration) control appear, but the prior art has the following significant drawbacks: The control dimension is single, namely, the on/off action is triggered only by the threshold value of the electroencephalogram signal at present, and the fine grading control of the grabbing force (such as light grabbing and tight grabbing) cannot be realized. The characteristic extraction is unstable, namely the forehead lobe brain electrical signal (EEG) has obvious non-stationarity and is extremely easy to be interfered by the Electrooculogram (EOG) and the myoelectricity (EMG), and the stable intention characteristic is difficult to be extracted in a short time by the traditional time-frequency analysis method. The existing brain control equipment is mainly controlled in an open loop mode, namely, after a brain sends out an instruction, the mechanical claw outputs the instruction with fixed voltage. When the object slides under the influence of gravity or inertia, the system cannot sense and adjust the grabbing force in real time, so that the grabbing success rate is low and the user experience is poor. Disclosure of Invention The embodiment of the application aims to provide a grabbing method, a device, equipment and a medium based on movement intention and concentration, which are used for solving the problems that the existing grabbing control method of a baby machine cannot finely control the grabbing force in a grading manner, the intention identification precision is low, an anti-drop mechanism is absent, and the user experience is low. In order to solve the above-mentioned problems, a first aspect provides a capturing method based on exercise intention and concentration, comprising: Step 100, constructing a support vector machine model, setting a plurality of acquisition electrodes in the brain of a user, acquiring electroencephalogram signals of the user at different grabbing intention levels according to a specified sampling rate, determining corresponding grabbing force levels, acquiring a training sample set of the intention levels and the grabbing force levels, and training the support vector machine model by using the training sample set to obtain a target support vector machine model; Step 200, collecting a real-time electroencephalogram signal of a corresponding channel electrode by using the collecting electrode, preprocessing the real-time electroencephalogram signal to obtain a preprocessed electroencephalogram signal, decomposing the preprocessed electroencephalogram signal by using an empirical mode decomposition algorithm to obtain a plurality of intrinsic mode signal components, calculating by using the plurality of intrinsic mode signal components to obtain a first energy sign sequence of a current channel, respectively carrying out feature extraction on the preprocessed electroencephalogram signal and the plurality of intrinsic mode signal components to obtain a first original feature sequence and a first intrinsic mode feature sequence, respectively carrying out standardization processing on the first energy sign sequence, the first original feature sequence and the first intrinsic mode feature sequence by using a Z-Score algorithm to obtain a second energy sign sequence, a second original feature sequence and a plurality of second intrinsic mode feature sequences, wherein the preprocessed electroencephalogram signal comprises Frequency band(s),Frequency bandFrequency band; Step 300, constructing an input feature vector by using the second energy bit sequence, the second original feature sequence and a plurality of second eigenvalue feature sequences, inputting the target support vector machine model for classification decision to obtain an intention level, and simultaneously obtaining the preprocessed electroencephalogram signals Frequency band signal power according to theAnd acquiring a user concentration degree average value by the frequency band signal power, and determining target output grabbing force according to the user concentration degree average value and the intention level. Step 400, detecting the stress state of the grabbing target in real time, performing PID operation to obtain the grabbing force difference between the actual output grabbing force and the target output grabbing force in real time, and if the grabbing target is detected to slide or the gra