CN112965595B - Modeling and predicting method for nerve electrical stimulation simulation touch sense - Google Patents
Modeling and predicting method for nerve electrical stimulation simulation touch sense Download PDFInfo
- Publication number
- CN112965595B CN112965595B CN202110219148.4A CN202110219148A CN112965595B CN 112965595 B CN112965595 B CN 112965595B CN 202110219148 A CN202110219148 A CN 202110219148A CN 112965595 B CN112965595 B CN 112965595B
- Authority
- CN
- China
- Prior art keywords
- nerve
- electrical stimulation
- modeling
- action
- force
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/016—Input arrangements with force or tactile feedback as computer generated output to the user
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Human Computer Interaction (AREA)
- Biomedical Technology (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Geometry (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
A modeling and prediction method of nerve electrical stimulation simulated touch, comprising the steps of: s1, measuring the force of muscles of a subject due to nerve electrical stimulation by applying nerve electrical stimulation pulse signals to the arms and/or hands of the subject; s2, analyzing and processing according to the multipath nerve electric stimulation pulse signals and the related data sets of the measured force magnitudes, and modeling a correlation model for input and output information of nerve electric stimulation simulation touch sense by utilizing the related data sets so as to predict the nerve electric stimulation simulation touch sense through the established model, wherein modeling the correlation model comprises establishing a touch sense simulation generation strategy through a Q-Learning algorithm or Deep Q-Learning algorithm. The application can effectively establish a model of the relevant factors of the nerve electrical stimulation simulation touch sense.
Description
Technical Field
The application relates to a modeling and predicting method for simulating touch sense by nerve electrical stimulation.
Background
Simulation of touch using a neural electrical stimulation method is an emerging topic in the scientific research field, and the existing research is mainly to acquire and analyze subjective feelings of a subject after receiving the neural electrical stimulation from a qualitative analysis perspective, and lacks related methods and devices for simulating touch using the neural electrical stimulation method, and methods for modeling and predicting input and output information of the neural electrical stimulation simulation touch using a related data set. How to effectively collect and analyze the data set of the simulation touch experiment of the large-scale nerve electric stimulation method is a problem to be solved in the prior art.
It should be noted that the information disclosed in the above background section is only for understanding the background of the application and thus may include information that does not form the prior art that is already known to those of ordinary skill in the art.
Disclosure of Invention
The application aims to overcome the defects of the background technology and provide a modeling and predicting method for simulating touch sense by nerve electric stimulation.
In order to achieve the above purpose, the present application adopts the following technical scheme:
a modeling and prediction method of nerve electrical stimulation simulated touch, comprising the steps of:
s1, measuring the force of muscles of a subject due to nerve electrical stimulation by applying nerve electrical stimulation pulse signals to the arms and/or hands of the subject;
s2, analyzing and processing according to the multipath nerve electric stimulation pulse signals and the related data sets of the measured force magnitudes, and modeling a correlation model for input and output information of nerve electric stimulation simulation touch sense by utilizing the related data sets so as to predict the nerve electric stimulation simulation touch sense through the established model, wherein modeling the correlation model comprises establishing a touch sense simulation generation strategy through a Q-Learning algorithm or Deep Q-Learning algorithm.
Further:
the method for establishing the haptic simulation generation strategy through the Q-Learning algorithm specifically comprises the following steps:
a1: the rewards reward which can be brought by the Q-Table to execute various actions under various states are stored in a Table; constructing a Q-Table by using the existing data set;
wherein, finite state set S is used, and finite action set A, transition ModelT (S, S, S') -P are used r (S ' |s, a) predicting the next state S ' based on the current state S and action a, i.e. the probability of taking action a to transition from S ' to S ', after which the state set S becomes S '; where s represents a specific state tester variable characteristic, such as heart rate and blood pressure, and a represents a specific action nerve electrical stimulation characteristic, such as frequency, waveform and intensity; reward R (s, a) represents the instant Reward Q-value: deltaF after the agent takes some action 2 ＝(F-F N ) 2 Wherein F is the actual force perceived by the tester, F N To be theoretically perceived by a testerThe amount of force to be applied;
a2: when the subject is in a certain state s, selecting an action a according to the value of the Q-Table, namely selecting a certain specific action nerve electric stimulation characteristic, wherein the reward obtained from the Table is Q (s, a), and the reward at the moment is the reward expected to be obtained; when action a is performed and the next state s' is transferred, an instant rewind can be obtained and recorded as r; the actual re-wait, denoted Q' (s, a), includes the immediate re-wait and future desired re-wait, and the future re-wait sets the discount factor γ, then the actual re-wait is expressed as follows:
a3: after obtaining the actual and expected re-word, using supervised learning to find the error between them and then updating, the updated value is the original Q (s, a), and the updating rule is as follows:
Q(s，a)＝Q(s，a)+α(Q′(s，a)-Q(s，a))
wherein α represents the learning rate of the gradient descent rule;
a4: the Q-Table optimization model for each tester is obtained by continuously updating the Q-Table, so that the establishment of the haptic simulation generation strategy is realized.
The value of γ is set to between 0 and 1, with 0 indicating that only immediate rewards are of interest and 1 indicating that future expected rewards are as important as immediate rewards.
The haptic simulation generation strategy is established through the Deep Q-Learning algorithm, and specifically comprises the following steps: compared with the Q-Learning algorithm, which has Experience Replay parts, the method repeatedly iterates and stores acquired samples, wherein each step is one sample, and each sample is a quadruple, and the method comprises the following steps: the Q value of various finite action actions of the current state, the immediate return obtained by the current finite action, and the Q value of various finite action actions of the next state; after the samples are obtained, the network is updated according to the Q-Learning update algorithm, where back propagation is required.
The measuring of the force of the muscle of the subject due to the nerve electrical stimulation maintains the sampling speed exceeding 6Hz and takes the dynamometer data in real time.
The nerve electric stimulation signals are pulse signals with the pulse frequency of not more than 500Hz, not more than 10mA and the duty ratio of 0.4-0.8, and not less than two paths of nerve electric stimulation signals and not more than ten paths of nerve electric stimulation signals are adopted.
Three paths of nerve electric stimulation signals are adopted, wherein two paths of nerve electric stimulation signals respectively act on the forearm, the other path of nerve electric stimulation signals act on the back of the hand, two pairs of electrode patches of the forearm are in an aligned posture, act on two large muscle bundles of triceps brachii, each pair of electrode patches is arranged up and down and act on the positions of muscle heads and muscle middles, and the positions are approximately four fifths and one third of the positions from the wrist to the elbow; the circle centers of a pair of patches on the back of the hand are positioned at two points along the line extending from the seam of the index finger, the middle finger and the ring finger and the line perpendicular to the line extending from the seam of the little finger and passing through the tiger mouth of the thumb.
The method comprises the steps of analyzing and processing according to the multipath nerve electric stimulation pulse signals and the related data sets of the measured force, and determining the force generated by the nerve electric stimulation pulse signals under the condition of different human body indexes, so that factors related to the generated force are determined from various factors, wherein the various factors comprise the frequency and the intensity of the nerve electric stimulation signals or various indexes of a subject.
A computer readable storage medium storing a computer program, characterized in that said computer program, when run by a processor, implements step S2 of said modeling and prediction method.
The application has the beneficial effects that:
by utilizing the modeling and prediction method of the nerve electric stimulation simulation touch sense, the data of the nerve electric stimulation simulation touch sense can be collected, analyzed and processed, and a model of relevant factors, particularly the magnitude of force, of the nerve electric stimulation simulation touch sense can be established. Has good application prospect in the aspect of nerve electrical stimulation simulation touch experiments.
Further, on the basis of the acquired nerve electrical stimulation simulation touch related data set, a touch simulation generation strategy is established by using a reinforcement Learning algorithm including Q-Learning and Deep Q-Learning. The reinforcement learning algorithm is combined with the nerve electric stimulation simulation touch data acquisition, so that quantitative simulation of force in the nerve electric stimulation simulation touch is realized.
The method has good application prospect in modeling and predicting the input and output information of the nerve electric stimulation simulation touch by using the related data set.
Drawings
FIG. 1 is a flow chart of a method of modeling and predicting neural electrical stimulation simulated haptics according to one embodiment of the present application;
FIG. 2 is a block diagram of a data acquisition and analysis device for a method of modeling and predicting neural-electric-stimulation simulated haptic sensations in accordance with one embodiment of the present application.
Detailed Description
The following describes embodiments of the present application in detail. It should be emphasized that the following description is merely exemplary in nature and is in no way intended to limit the scope of the application or its applications.
It will be understood that when an element is referred to as being "mounted" or "disposed" on another element, it can be directly on the other element or be indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or be indirectly connected to the other element. In addition, the connection may be for both a fixing action and a coupling or communication action.
It is to be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like are merely for convenience in describing embodiments of the application and to simplify the description by referring to the figures, rather than to indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus are not to be construed as limiting the application.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the embodiments of the present application, the meaning of "plurality" is two or more, unless explicitly defined otherwise.
The embodiment of the application provides a modeling and predicting method for simulating touch sense by nerve electrical stimulation, which comprises the following steps:
s1, measuring the force of muscles of a subject due to nerve electrical stimulation by applying nerve electrical stimulation pulse signals to the arms and/or hands of the subject;
s2, analyzing and processing according to the multipath nerve electric stimulation pulse signals and the related data sets of the measured force magnitudes, and modeling a correlation model for input and output information of nerve electric stimulation simulation touch sense by utilizing the related data sets so as to predict the nerve electric stimulation simulation touch sense through the established model, wherein modeling the correlation model comprises establishing a touch sense simulation generation strategy through a Q-Learning algorithm or Deep Q-Learning algorithm.
The modeling and prediction method of the specific embodiment may be implemented by a data acquisition and analysis apparatus as shown in fig. 2. The nerve electric stimulation simulation tactile data acquisition and analysis device comprises a processing device, a multipath nerve electric stimulation pulse signal generator, electrodes and a dynamometer, wherein the processing device controls nerve electric stimulation pulse signals output by the multipath nerve electric stimulation pulse signal generator to act on the arm and/or the hand of a subject through a plurality of electrodes so as to stimulate muscle nerves of the subject, the dynamometer is used for measuring the force magnitude of muscles of the subject due to nerve electric stimulation, the measured force data are sent to the processing device, the processing device analyzes and processes according to the multipath nerve electric stimulation pulse signals and the measured force magnitude related data sets, and the related data sets are used for modeling a correlation model for input and output information of the nerve electric stimulation simulation tactile, so that the nerve electric stimulation simulation tactile prediction is performed through the established model.
Referring to fig. 2, in a preferred embodiment, the data acquisition and analysis device further comprises a height gauge for adjusting the load cell to a state of just abutting against a measurement site of a subject without generating a force during measurement of the load cell.
In a preferred embodiment, the load cell includes a load cell sensor and a load cell display.
The processing device determines the magnitude of the force generated by the nerve electric stimulation pulse signal under the condition of different human body indexes through analyzing and processing the signals, thereby determining factors related to the magnitude of the generated force from various factors, wherein the various factors comprise the frequency, the intensity of the nerve electric stimulation signal or various indexes of a subject.
Specific embodiments of the present application are described further below.
Referring to fig. 1 and 2, in one embodiment, to obtain feedback from a subject regarding the nerve electrical stimulation signal, a nerve electrical stimulation pulse signal is input to the arm and/or hand of the subject to stimulate the muscle nerve of the subject, and the magnitude of the force generated by this stimulation is output, thereby establishing a model of the correlation between these two types of data regarding the nerve electrical stimulation pulse signal and the magnitude of the force.
The processing device determines the magnitude of the generated force of the nerve electric stimulation signal under the condition of different human indexes through processing the signals, so as to determine factors (such as the frequency and the intensity of the nerve electric stimulation signal or various indexes of a subject) with larger relevance with touch sense (particularly the magnitude of the force), or establish a more complex relevant model.
For acquisition and analysis of the relevant data set, in a preferred embodiment, the neural electrical stimulation signal is selected to be a pulse signal of no more than 500Hz, no more than 10mA, and a duty cycle of 0.4-0.8.
Preferably, at least two nerve electric stimulation signals and at most ten nerve electric stimulation signals are adopted, and materials such as circular silica gel patches with diameters of not more than 4cm are respectively acted on corresponding areas of arms (including a big arm, a small arm and a palm).
In a preferred embodiment, three nerve electrical stimulation signals are used, two of which act on the forearm and the other on the back of the hand. The two pairs of patches of the forearm are in an aligned posture and act on two large muscle bundles of the triceps brachii, each pair is arranged up and down and acts on the muscle head and the middle section of the muscle, and the positions approximate to the positions from the wrist to the quarter and the third of the elbow. The circle centers of a pair of patches on the back of the hand are positioned at two points along the line extending from the seam of the index finger, the middle finger and the ring finger and the line perpendicular to the line extending from the seam of the little finger and passing through the tiger mouth of the thumb.
With no nerve electrical stimulation at all as a calibration point, the load cell preferably maintains a sampling rate in excess of 6Hz after application of the nerve electrical stimulation signal, and load cell data is taken in real time.
Using the correlated data set, in combination with the neuro-electrical stimulation signals and the load cell data, the processing device may perform a cross-over analysis of correlated factors including subject indices and the like to determine a data model.
In a preferred embodiment, in step S2, a haptic simulation generation strategy is established for the acquired dataset simulating the haptic sensation using neural electrical stimulation using a reinforcement Learning algorithm including Q-Learning and Deep Q-Learning. The specific treatment method comprises the following steps:
method A: for smaller-capacity data sets, the establishment of the haptic simulation generation strategy is mainly performed through a Q-Learning algorithm.
A1: the Q-Learning algorithm uses a Table to store the reward that the Q-Table can bring about in each state, and Table 1 below shows two states s1, s2, with two actions a1, a2 in each state, with the values in the Table representing reward.
TABLE 1
| reward | action 1 | action 2 |
| state 1 | reward 1 | reward 2 |
| state 2 | reward 3 | reward 4 |
Each value in table 1 is defined as Q (s, a) and represents the reward obtained by performing action a in state s, then a greedy approach can be used in selecting which action is most valuable to perform.
The method comprises the following steps:
s: representing a finite state set, wherein s represents a certain state tester variable characteristic (heart rate, blood pressure two characteristics);
a: a finite action set, a represents the electrical stimulation characteristics (frequency, waveform, intensity, etc.) of a specific action nerve;
Transition ModelT(S,s,S′)～P r (s' |s, a): predicting the next state S ' according to the current state S and the action a, namely taking the probability of the action a to be transferred from S, and changing the state collection S into S ' after the action a to be transferred to S '; reward R (s, a) represents the instant prize Q-value: ΔF after the agent takes some action 2 ＝(F-F N ) 2 Wherein F is the actual force perceived by the tester, F N To the magnitude of the simulated force that would theoretically be perceived by the tester.
The Q-Table is constructed using the existing data set as described above.
A2: when this is tested in a state s, action a is selected according to the value of Q-Table, i.e. a particular action neural stimulation feature is selected, and the reward that can be obtained from the Table is Q (s, a), where the reward is not actually obtained but is expected to be obtained, but is not actually obtained. When action a is performed and the next state s ' is transferred, an immediate re-ward (denoted as r) can be obtained, but in addition to the immediate re-ward, the transferred state s ' is considered for future desired re-ward, so the actual re-ward (denoted as Q ' (s, a)) consists of two parts: instant and future expected reward, and future reward is often uncertain, for which reason the discount factor γ is set, the true reward is expressed as follows:
the value of γ is preferably set between 0 and 1, with 0 indicating that only immediate payback is of concern and 1 indicating that future expected payback is as important as immediate payback.
A3: after obtaining the actual and expected reward, using supervised learning to find the error between them and then updating, the updated value is the original Q (s, a), and the updating rule is as follows:
Q(s，a)＝Q(s，a)+α(Q′(s，a)-Q(s，a))
wherein α represents the learning rate of the gradient descent rule;
a4: the Q-Table optimization model for each tester is obtained by continuously updating the Q-Table, so that the establishment of the haptic simulation generation strategy is realized.
Method B: and establishing a tactile simulation generation strategy mainly through Deep Q-Learning algorithm aiming at a complete and large-capacity data set.
The method B and the method A are mostly coincident in the specific implementation mode, and the main algorithm flow is as follows:
the most difference between the DQL algorithm and Q-Learning is that Experience Replay parts are added, the mechanism is that repeated iteration is first performed, samples obtained in the iteration steps are stored, each step is a sample, each sample is a quadruple, and the method comprises the following steps: the Q value of each action in the current state, the instant return obtained by the currently adopted action, and the Q value of each action in the next state. After obtaining a sample, the network can be updated according to the Q-Learning update algorithm mentioned above, except that back propagation is required at this time.
The background section of the present application may contain background information about the problems or environments of the present application and is not necessarily descriptive of the prior art. Accordingly, inclusion in the background section is not an admission of prior art by the applicant.
The foregoing is a further detailed description of the application in connection with specific/preferred embodiments, and it is not intended that the application be limited to such description. It will be apparent to those skilled in the art that several alternatives or modifications can be made to the described embodiments without departing from the spirit of the application, and these alternatives or modifications should be considered to be within the scope of the application. In the description of the present specification, reference to the terms "one embodiment," "some embodiments," "preferred embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Those skilled in the art may combine and combine the features of the different embodiments or examples described in this specification and of the different embodiments or examples without contradiction. Although embodiments of the present application and their advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the scope of the application as defined by the appended claims.
Claims (7)
1. A method for modeling and predicting simulated haptics by nerve electrical stimulation, comprising the steps of:
s1, measuring the force of muscles of a subject due to nerve electrical stimulation by applying nerve electrical stimulation pulse signals to the arms and/or hands of the subject;
s2, analyzing and processing according to a multi-path nerve electric stimulation pulse signal and a related data set of measured force magnitude, and modeling a correlation model of input and output information of nerve electric stimulation simulation touch sense by utilizing the related data set so as to predict the nerve electric stimulation simulation touch sense through the established model, wherein modeling the correlation model comprises establishing a touch sense simulation generation strategy through a Q-Learning algorithm or Deep Q-Learning algorithm;
for a small-capacity data set, establishing a haptic simulation generation strategy through a Q-Learning algorithm, wherein the method specifically comprises the following steps of:
a1: using a Table Q-Table to store rewards review which can be brought by executing various actions in various states; each value in the table is defined as Q (s, a) representing the reward obtained by performing action a in state s; constructing a Q-Table by using the existing data set; wherein S represents a finite state set, wherein S represents a subject variable characteristic of a particular state, the variable characteristic being heart rate, voltage; a represents a finite action set, wherein a represents a specific action neural electrical stimulation characteristic, which is frequency, waveform, intensity; transition Model T (S, S, S') -P r (S '|s, a) represents the probability of predicting the next state S' from the current state S and the action a, i.e., taking action a to transition from S to S ', after which the state aggregate S becomes S'; reward R (s, a) represents the instant prize Q-value: ΔF after the agent takes some action 2 ＝(F-F N ) 2 Wherein F is the actual force perceived by the tester, F N To the magnitude of the simulated force that would theoretically be perceived by the tester；
A2: when the subject is in a certain state s, selecting an action a according to the value of the Q-Table, namely selecting a certain specific action nerve electric stimulation characteristic, wherein the reward obtained from the Table is Q (s, a), and the reward at the moment is the reward expected to be obtained; when action a is performed and the next state s' is transferred, an instant rewind can be obtained and recorded as r; the actual re-wait, denoted Q' (s, a), includes the instant re-wait and the re-wait expected to be acquired, and the re-wait expected to be acquired is uncertain, so the discount factor γ is set, and the actual re-wait is expressed as follows:
a3: after obtaining the actual and expected re-word, using supervised learning to find the error between them and then updating, the updated value is the original Q (s, a), and the updating rule is as follows:
Q(s,α)＝Q(s,α)+α(Q'(s,α)-Q(s,α))
wherein α represents the learning rate of the gradient descent rule;
a4: the Q-Table is continuously updated to obtain a Q-Table optimization model aiming at each tester, so that the establishment of a haptic simulation generation strategy is realized;
for a large-capacity data set, establishing a tactile simulation generation strategy through a Deep Q-Learning algorithm, wherein the method specifically comprises the following steps of: the Deep Q-Learning algorithm is further Experience Replay than the Q-Learning algorithm, and the iteration is repeated, and samples obtained in the iteration steps are stored, where each step is a sample, and each sample is a quadruple, including: the Q value of various finite action actions of the current state, the immediate return obtained by the current finite action, and the Q value of various finite action actions of the next state; after obtaining a sample, the network is updated according to the Q-Learning update algorithm described above.
2. The modeling and prediction method of claim 1, wherein the value of γ is set to between 0 and 1, and setting 0 indicates that only immediate re-ward is of interest, and setting 1 indicates that re-ward expected to be acquired is as important as immediate re-ward.
3. Modeling and prediction method according to any of the claims 1-2, characterized in that the load cell data is taken in real time maintaining a sampling rate exceeding 6Hz while measuring the magnitude of the force of the muscles of the subject due to the nerve electrical stimulation.
4. The modeling and prediction method of any of claims 1-2, wherein the nerve electrical stimulation signals are pulse signals of no more than 500Hz, no more than 10mA, and a duty cycle of 0.4-0.8, and no less than two and no more than ten nerve electrical stimulation signals are used.
5. Modeling and prediction method according to any of the claims 1 to 2, characterized in that three electrical nerve stimulation signals are used, two of which act on the forearm and the other on the dorsum of the hand respectively, the two pairs of electrode patches of the forearm being in a parallel posture, acting on the two large muscle bundles of the triceps brachii, each pair being arranged up and down, acting on the muscle head and the middle section of the muscle, approximately four fifths and one third of the way from the wrist to the elbow; the circle centers of a pair of patches on the back of the hand are positioned at two points along the line extending from the seam of the index finger, the middle finger and the ring finger and the line perpendicular to the line extending from the seam of the little finger and passing through the tiger mouth of the thumb.
6. A modeling and prediction method as claimed in any of claims 1 to 2 wherein the magnitude of the force generated by the nerve electrical stimulation pulse signal in the case of different body metrics is determined by analysis and processing of the data sets relating the magnitude of the measured force to the nerve electrical stimulation pulse signal, whereby factors relating to the magnitude of the generated force are determined from a plurality of factors including the frequency, intensity of the nerve electrical stimulation signal, or various metrics of the subject.
7. A computer readable storage medium storing a computer program, characterized in that the computer program, when run by a processor, implements step S2 of the modeling and prediction method according to any of the claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110219148.4A CN112965595B (en) | 2021-02-26 | 2021-02-26 | Modeling and predicting method for nerve electrical stimulation simulation touch sense |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110219148.4A CN112965595B (en) | 2021-02-26 | 2021-02-26 | Modeling and predicting method for nerve electrical stimulation simulation touch sense |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112965595A CN112965595A (en) | 2021-06-15 |
| CN112965595B true CN112965595B (en) | 2023-09-05 |
Family
ID=76275861
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110219148.4A Active CN112965595B (en) | 2021-02-26 | 2021-02-26 | Modeling and predicting method for nerve electrical stimulation simulation touch sense |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112965595B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117958764A (en) * | 2024-04-01 | 2024-05-03 | 北京科技大学 | Method and system for predicting stable working current of percutaneous electrical stimulation of BP neural network |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20150062584A (en) * | 2013-11-29 | 2015-06-08 | 숭실대학교산학협력단 | Feedback brain stimuation device controlling in periodic and non-periodic mode |
| CN106681149A (en) * | 2017-01-11 | 2017-05-17 | 浙江大学 | Bumblebee robot abdomen swinging control method based on virtual reality and reinforcement learning |
| CN109688990A (en) * | 2016-09-06 | 2019-04-26 | 新感知公司 | For providing a user the method and system of attached sensory information |
| CN111659010A (en) * | 2020-07-16 | 2020-09-15 | 郑州大学 | Limb functional electrical stimulation control method and limb rehabilitation robot |
| CN111722709A (en) * | 2020-05-15 | 2020-09-29 | 华南理工大学 | Virtual force feedback interaction system and method based on myoelectric evaluation |
| CN111795761A (en) * | 2020-06-24 | 2020-10-20 | 菲尼克斯(上海)环境控制技术有限公司 | Method for predicting cabinet inlet air temperature based on reinforcement learning model |
-
2021
- 2021-02-26 CN CN202110219148.4A patent/CN112965595B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20150062584A (en) * | 2013-11-29 | 2015-06-08 | 숭실대학교산학협력단 | Feedback brain stimuation device controlling in periodic and non-periodic mode |
| CN109688990A (en) * | 2016-09-06 | 2019-04-26 | 新感知公司 | For providing a user the method and system of attached sensory information |
| CN106681149A (en) * | 2017-01-11 | 2017-05-17 | 浙江大学 | Bumblebee robot abdomen swinging control method based on virtual reality and reinforcement learning |
| CN111722709A (en) * | 2020-05-15 | 2020-09-29 | 华南理工大学 | Virtual force feedback interaction system and method based on myoelectric evaluation |
| CN111795761A (en) * | 2020-06-24 | 2020-10-20 | 菲尼克斯(上海)环境控制技术有限公司 | Method for predicting cabinet inlet air temperature based on reinforcement learning model |
| CN111659010A (en) * | 2020-07-16 | 2020-09-15 | 郑州大学 | Limb functional electrical stimulation control method and limb rehabilitation robot |
Non-Patent Citations (1)
| Title |
|---|
| Playing Atari with Deep Reinforcement Learning;Volodymyr Mnih;《ArXiv》;1-9 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112965595A (en) | 2021-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Macdonald | Reflections on the history of electrochemical impedance spectroscopy | |
| KR102008196B1 (en) | Apparatus for generating of Potassium level determination model using ECG and method thereof | |
| Ma et al. | An optimal electrical impedance tomography drive pattern for human-computer interaction applications | |
| Luo et al. | Possibility of quantifying TCM finger-reading sensations: I. Bi-sensing pulse diagnosis instrument | |
| CN109875565A (en) | A kind of cerebral apoplexy upper extremity exercise function method for automatically evaluating based on deep learning | |
| González-Izal et al. | sEMG wavelet-based indices predicts muscle power loss during dynamic contractions | |
| TW201116318A (en) | System and method for analyzing muscle performance status at swing action, and program products thereof | |
| CN112965595B (en) | Modeling and predicting method for nerve electrical stimulation simulation touch sense | |
| CN108209912A (en) | A kind of electromyographic signal collection method and device | |
| Mylott et al. | Bioelectrical impedance analysis as a laboratory activity: At the interface of physics and the body | |
| CN112915384B (en) | Data acquisition and analysis device for nerve electrical stimulation simulation touch sense | |
| CN109919099A (en) | A kind of user experience evaluation method and system based on Expression Recognition | |
| JP5431147B2 (en) | Impedance measurement | |
| CN105354414A (en) | Hierarchical analysis-based human body health condition assessment method | |
| CN101889874B (en) | Evaluation method of ergonomics design of ultrasonic probe | |
| AU2019212111B2 (en) | Methods for establishing the stiffness of a bone using mechanical response tissue analysis | |
| CN115565685A (en) | Electrode-skin impedance model parameter optimization method based on mixed group intelligent algorithm | |
| CN107242877B (en) | Evaluation system and evaluation method for emotion stability sensing system training | |
| Sun et al. | Improved Hill-type musculotendon models with activation-force-length coupling | |
| Koryahin et al. | Didactic opportunities of information-communication Technologies in the Control of physical education | |
| CN103300854B (en) | Human acupuncture point chaotic characteristic measuring instrument based on virtual instrument and measuring method thereof | |
| Samarpita et al. | Mental Stress Classification from Brain Signals using MLP Classifier | |
| CN115983037B (en) | Myoelectricity and optimized coupling muscle force calculation method based on muscle cooperative constraint | |
| EP4032478A1 (en) | Characterizing soft tissue stress for ameliorating injury in performing a process | |
| Jung et al. | Surface Electromyogram Simulator for Myoelectric Prosthesis Testing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |