CN107622260A - Lower limb gait phase identification method based on multi-source bio signal - Google Patents
Lower limb gait phase identification method based on multi-source bio signal Download PDFInfo
- Publication number
- CN107622260A CN107622260A CN201711013763.XA CN201711013763A CN107622260A CN 107622260 A CN107622260 A CN 107622260A CN 201711013763 A CN201711013763 A CN 201711013763A CN 107622260 A CN107622260 A CN 107622260A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- particle
- msubsup
- multiple dimensioned
- 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.)
- Pending
Links
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The present invention proposes a kind of lower limb gait phase identification method based on multi-source bio signal.The present invention obtains human body lower limbs surface electromyogram signal, plantar pressure characteristic value and knee joint angle characteristic value first.Secondly, sEMG signals are subjected to WAVELET PACKET DECOMPOSITION and extract multiple dimensioned energy and multiple dimensioned fuzzy entropy feature；Then, one group of characteristic vector is formed with plantar pressure characteristic value and knee joint energy eigenvalue using after principal component analysis (PCA) method dimension-reduction treatment to the sEMG signal characteristics value of extraction.Finally, characteristic vector is inputted into particle group optimizing least square method supporting vector machine (PSO LSSVM) realization mutually to identify human body lower limbs gait.The present invention, which closes, can be achieved the mutually high discrimination of human body lower limbs gait, and recognizer can be used for designing various recovering aid equipment, intelligent artificial limb, walk supporting device etc..
Description
Technical field
The invention belongs to area of pattern recognition, is related to one kind and is based on surface electromyogram signal (sEMG), knee joint angle and foot
The human body lower limbs gait facies model recognition methods of base pressure force signal, lower limb gait are mutually included after supporting early support mid-term, support
Phase, swing early stage, swing the later stage.
Background technology
Human body lower limbs gait Recognition technology has a wide range of applications, and because of the particularity of lower limb action message, gets over
Put into come more scholars in the research of lower limb Gait Recognition.The technology can be used in medical monitoring, recovering aid is treated, machine
A series of field of man-machine interaction problems of design such as device people, motion prediction.It can assess whether patient deposits by gait analysis
In abnormal gait and the nature and extent of abnormal gait.Sang Wan Lee etc. are by gathering leg surface myoelectric (Surface
Electromyography, sEMG) signal Input Fuzzy Neural Network realizes identification to lower limb gait.Faisal Ahmed
Realize that the three-dimensional gait of human body identifies Deng using Kinect sensor.Liu F etc. capture human body row by new side view drawing method
Joint angles data when walking, by training stage and cognitive phase, realize the identification to human body lower limbs gait.
Gait phase identifying system based on multi-source bio signal merits attention.Due to measuring the progress with information technology,
It has been recognized that sEMG signals have huge potentiality in terms of man-machine interaction, because the sEMG letters that contraction of muscle is supervened
It number can reflect the motion intention and state of human body.Plantar pressure information can intuitively reflect the current lower limb gait phase of human body
State.Knee joint angle generating period can change in different gait phases.The fusion sEMG such as high cloud circle signals, plantar pressure
Lower limb motion mode is identified with thigh and calf knee joint angle, realizes effective control of above-knee prosthesis.Liu Lei etc. uses sEMG
Signal, plantar pressure and Hip Angle multi-source bio signal realize the correct identification to lower limb motion mode more than 90%
Rate.
The content of the invention
The present invention is to realize that human body lower limbs gait mutually identifies and further improved the accuracy of identification, it is proposed that Yi Zhongrong
Close surface electromyogram signal (sEMG), knee joint angle mutually identifies with the human body lower limbs gait of plantar pressure signal.First, by sEMG
Signal carries out WAVELET PACKET DECOMPOSITION and extracts multiple dimensioned energy and multiple dimensioned fuzzy entropy feature；Then, to the sEMG signal characteristics of extraction
Value forms one group using after principal component analysis (PCA) method dimension-reduction treatment with plantar pressure characteristic value and knee joint energy eigenvalue
Characteristic vector.Finally, characteristic vector is inputted into particle group optimizing least square method supporting vector machine (PSO-LSSVM) model classifiers
Realize that carrying out gait to human body lower limbs movable information mutually identifies.
In order to realize the above object the inventive method mainly includes the following steps that：
Step 1. obtains human body lower limbs surface electromyogram signal, plantar pressure characteristic value and knee joint angle characteristic value.Specifically
Process is as follows：
Myoelectricity collection is pasted respectively in right lateral thigh vastus medialis, rectus femoris, long adductor muscle and the tensor fasciae late muscle of subject
Sensor, electromyographic signal is obtained by myoelectricity acquisition system；Pressed by obtaining vola installed in the plantar pressure sensor of sole
Power characteristic value；Acceleration transducer is placed respectively in the small leg outer side of thigh, and lower limb difference appearance is measured from double-axel acceleration sensor
Acceleration axial component during state calculates knee joint angle characteristic value.
The surface electromyogram signal that step 2. obtains step 1 carries out five layers of WAVELET PACKET DECOMPOSITION, by nine low frequencies after decomposition
Subspace surface electromyogram signal reconstruct, the multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned fuzzy entropy after then extraction reconstructs
Value.
The multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned fuzzy entropy that step 3. obtains step 2 use principal component
Analysis method dimension-reduction treatment.Detailed process is as follows：The multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned obtained from step (2)
Fuzzy entropy 72 is tieed up totally, using principal component analytical method, according to the principle of gait phase character storage rate more than 95%, from 72 Wei Te
Selection can summarize the 10 dimensional features composition myoelectricity characteristic vector of primitive character value in sign.
Step 3 is obtained myoelectricity characteristic vector and forms one with plantar pressure characteristic value and knee joint energy eigenvalue by step 4.
12 dimensional feature vectors of group.
Step 5. enters the 12 dimensional feature vectors input particle group optimizing least square method supporting vector machine that step (4) is obtained
Row gait mutually identifies.
Described particle group optimizing least square method supporting vector machine design is as follows：
The distribution characteristics of 12 dimensional feature vectors obtained according to step 4, a most young waiter in a wineshop or an inn is optimized by particle swarm optimization algorithm
Multiply the kernel functional parameter and penalty factor of SVMs.Concretely comprise the following steps：
(1) iterations, initial position, speed and the current location of particle, the position of optimal particle in population are initialized；
(2) sample is divided into training set and test set, its adaptation is determined according to least square method supporting vector machine categorised decision
Spend functional value；
(3) fitness of each particle is calculated, p is updated according to fitness sizeiAnd pg, and grain is updated according to following formula respectively
Sub- speed, particle position, inertia weight and Studying factors.
Z in formulai=(zi1,zi2,...,zid), vi=(vi1,vi2,...,vid) position and the speed of i-th particle are represented respectively
Degree, d is space dimensionality.K belongs to iterations, and w is inertia weight, r1、r2For two random functions, c1、c2For scholar's factor.pi
=(pi1,pi2,...,pid) it is the optimal location that particle search arrives；pg=(pg1,pg2,...pgd) for population search it is optimal
Position；
(4) if fitness meets desired output, optimizing terminates, and otherwise believes the particle rapidity after renewal, particle position
Breath substitutes into step 4- (3), until fitness meets desired output；；The iterative algorithm of population optimizing more than, is made
The optimal value of the minimum penalty factor of least square method supporting vector machine error and kernel function radius parameter.
It is of the invention compared with existing human body lower limbs gait phase identification method, there are following features：
The human body lower limbs gait based on sEMG signals, plantar pressure and knee joint angle proposed mutually identifies, by collection
SEMG signals WAVELET PACKET DECOMPOSITION extracts multiple dimensioned energy value and multiple dimensioned fuzzy entropy feature and passes through principal component analysis
(Principal component analysis, PCA) carries out dimension-reduction treatment, with knee joint angle and plantar pressure energy value group
Particle group optimizing (Particle swarm optimization, PSO) least square supporting vector is inputted into one group of characteristic vector
Machine (Least squares support vector machine, LSSVM) grader mutually identifies that LSSVM has to lower limb gait
There are strict Fundamentals of Mathematics and good generalization ability, it is possible to prevente effectively from local optimum problem and overcoming dimension disaster, particle
Swarm parameter optimization ensures the diversity of population, and search is quickly found out global LSSVM optimized parameters.It the method achieve human body lower limbs
The high discrimination of gait phase and the reliability for improving systematic function, so as to which institute's extracting method can apply to design various auxiliary health
Multiple equipment, walk supporting device, intelligent artificial limb etc..
Brief description of the drawings
Fig. 1 is the implementing procedure figure of the present invention；
Fig. 2 is that the lower limb gait information of the embodiment of the present invention gathers experimental system；
Fig. 3 is the principal component analysis Dimension Reduction Analysis of the embodiment of the present invention；
Fig. 4 is that the PSO of the embodiment of the present invention optimizes the selection of LSSVM model parameters.
Embodiment
Embodiments of the invention are elaborated below in conjunction with the accompanying drawings：The present embodiment using technical solution of the present invention before
Put and implemented, give detailed embodiment and specific operating process.
As shown in figure 1, the present embodiment comprises the following steps：
Step 1, obtain human body lower limbs surface electromyogram signal, plantar pressure characteristic value and knee joint angle characteristic value.Specifically
Process is as follows：
Three healthy students of experimental selection are subject, 2 boy students, the age (25 ± 5), height (170.0 ± 5.0)
Cm, body weight (65.0 ± 5.0) kg, 1 schoolgirl, at the age 23, height 163.0cm, body weight 45.0kg, numbering is respectively 1~3, real
Testing the last week does not have strenuous exercise.Distinguish in the right lateral thigh vastus medialis of subject, rectus femoris, long adductor muscle and tensor fasciae late muscle
Myoelectricity collection sensor is pasted, electromyographic signal is obtained by myoelectricity acquisition system；Sensed by the plantar pressure installed in sole
Device obtains plantar pressure characteristic value；Acceleration transducer is placed respectively in the small leg outer side of thigh, is surveyed from double-axel acceleration sensor
Acceleration axial component when measuring lower limb difference posture calculates knee joint angle characteristic value.In a gait cycle, collection
Early stage, mid-term and later stage are supported, swings 3 kinds of synchronizing signals in 5 stages of early stage and later stage, through PC data line transfers to calculating
Machine.Signal acquisition process is as shown in Figure 2.
The surface electromyogram signal that step 2 obtains step 1 carries out five layers of WAVELET PACKET DECOMPOSITION, and nine after decomposition are low
The reconstruct of frequency subspace surface electromyogram signal, the then multiple dimensioned energy value of surface electromyogram signal after extraction reconstruct and multiple dimensioned fuzzy
Entropy.
After five stage sEMG signal live parts in one gait cycle of interception and filtering process, extract each dynamic
The multiple dimensioned energy value of Zuo Mei roads sEMG signals and multiple dimensioned Sample Entropy, multiple dimensioned energy value and multiple dimensioned fuzzy entropy distribution such as table
1, shown in table 2.
The multiple dimensioned energy value of each gait phase of table 1
| Gait phase | Passage 1 | Passage 2 | Passage 3 | Passage 4 | Passage 5 | Passage 6 | Passage 7 | Passage 8 | Passage 9 |
| Support early stage | 22.6±0.3 | 23.0±0.3 | 18.1±0.3 | 17.6±0.2 | 19.6±0.3 | 18.3±0.3 | 18.6±0.2 | 21.6±0.3 | 19.3±0.3 |
| Support mid-term | 16.1±0.2 | 18.2±0.3 | 17.0±0.2 | 17.4±0.3 | 15.7±0.2 | 18.7±0.2 | 18.1±0.2 | 21.4±0.3 | 17.8±0.3 |
| Support the later stage | 17.0±0.3 | 18.3±0.3 | 19.2±0.3 | 18.3±0.3 | 20.1±0.3 | 19.6±0.3 | 19.3±0.3 | 20.7±0.3 | 19.0±0.3 |
| Swing early stage | 22.4±0.4 | 19.6±0.3 | 22.3±0.4 | 21.8±0.4 | 22.1±0.4 | 23.4±0.4 | 21.6±0.3 | 20.8±0.3 | 22.6±0.4 |
| Swing the later stage | 19.7±0.3 | 22.7±0.4 | 22.3±0.3 | 19.3±0.3 | 21.3±0.3 | 19.3±0.3 | 19.0±0.3 | 22.3±0.3 | 21.7±0.3 |
The multiple dimensioned fuzzy entropy of each gait phase of table 2
| Gait phase | Passage 1 | Passage 2 | Passage 3 | Passage 4 | Passage 5 | Passage 6 | Passage 7 | Passage 8 | Passage 9 |
| Support early stage | 1.30±0.02 | 1.32±0.03 | 1.82±0.03 | 1.03±0.03 | 0.64±0.03 | 0.43±0.02 | 0.72±0.03 | 2.12±0.02 | 1.80±0.03 |
| Support mid-term | 0.70±0.02 | 0.40±0.02 | 0.82±0.02 | 0.82±0.02 | 0.52±0.02 | 0.42±0.02 | 0.64±0.02 | 1.32±0.03 | 0.51±0.02 |
| Support the later stage | 1.57±0.03 | 0.32±0.02 | 1.14±0.03 | 0.91±0.02 | 1.21±0.03 | 1.02±0.03 | 0.32±0.02 | 0.44±0.02 | 0.32±0.02 |
| Swing early stage | 1.58±0.03 | 0.97±0.03 | 1.02±0.03 | 0.70±0.02 | 0.61±0.03 | 2.20±0.04 | 1.70±0.03 | 0.52±0.02 | 0.51±0.02 |
| Swing the later stage | 0.20±0.01 | 0.98±0.03 | 1.90±0.03 | 0.78±0.03 | 0.59±0.03 | 1.32±0.03 | 0.65±0.02 | 2.22±0.04 | 0.62±0.02 |
The multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned fuzzy entropy that step 3 obtains step 2 using it is main into
Analysis (PCA) method dimension-reduction treatment.It is specific as follows：The surface electromyogram signal of the four pieces of muscle obtained from step (2) is multiple dimensioned
Energy value and multiple dimensioned fuzzy entropy 72 are tieed up totally, using principal component analysis (PCA) method, according to gait phase character storage rate 95%
Principle above, the 10 dimensional features composition myoelectricity characteristic vector that can summarize primitive character value is selected from 72 dimensional features.
Feature extraction as stated above, by principal component analysis dimension-reduction algorithm dimension-reduction treatment, Fig. 3 gives Dimension Reduction Analysis,
It can be seen that five kinds of gait phase character storage rates are maximum when 10 tie up, i.e., contribution rate of accumulative total is more than 95%.Consider choosing
Primitive character value can be summarized by selecting 10 dimensions.
Step 3 is obtained myoelectricity characteristic vector and formed with plantar pressure characteristic value and knee joint energy eigenvalue by step 4
One group of 12 dimensional feature vector
12 dimensional feature vectors that step 5 is obtained step 4 input particle group optimizing least square method supporting vector machine
(PSO-LSSVM) model classifiers progress gait mutually identifies.
130 groups of selection is concentrated from each action data, 30 groups are used as training sample, and 100 groups are used as test sample, according to step
The distribution characteristics of rapid four 12 dimensional feature vectors obtained, passes through particle swarm optimization algorithm Optimized Least Square Support Vector
Kernel functional parameter and penalty factor.Fig. 4 is that PSO optimizes the selection of LSSVM model parameters, works as c=0.1, during g=5.8737, model
Parameter is optimal solution.
The lower limb multi-source information for testing collection is subjected to feature extraction, Fusion Features and pattern-recognition using the inventive method
Afterwards, as a result as shown in table 3.Believe with the Gait Recognition rate using only electromyographic signal and using only leg angle and plantar pressure
The Gait Recognition rate of breath is compared, and the lower limb gait phase identification method of Multi-source Information Fusion proposed by the present invention has higher identification
Rate.
In order that PSO-LSSVM prediction results have comparativity, carried out using genetic algorithm optimization LSSVM (GA-LSSVM)
Contrast test, enter data into LSSVM and be trained.Discrimination is as shown in table 4.Can from table 3 and the statistics of table 4
Go out in mutually being identified by the lower limb gait of two kinds of graders, PSO-LSSVM discrimination is higher than GA-LSSVM, and each stage reaches
More than 95%, illustrate that the gait phase average discrimination of three subjects is more stable, while illustrate that PSO-LSSVM can be more accurate
Really, punishment parameter and the optimum organization value of kernel functional parameter are more effectively found, identification error is accordingly reduced, while trains and divides
Class and recognition correct rate rise.Optimize LSSVM parameters using PSO, in solving the problems, such as small sample, non-linear and height mode identification
Show some superiority.As can be seen here, the multiple dimensioned energy of sEMG signals and multiple dimensioned fuzzy entropy feature, knee joint angle of extraction
With plantar pressure energy value and PSO-LSSVM classifiers combinations unfavorable factor can be overcome accurately to realize lower limb Gait Recognition.
Table 3 is compared using PSO-LSSVM grader gait phase recognition correct rates
Table 4 is compared using GA-LSSVM grader gait phase recognition correct rates
Claims (3)
1. the lower limb gait phase identification method based on multi-source bio signal, it is characterised in that this method comprises the following steps：
Step 1. obtains human body lower limbs surface electromyogram signal, plantar pressure characteristic value and knee joint angle characteristic value；
The surface electromyogram signal that step 2. obtains step 1 carries out five layers of WAVELET PACKET DECOMPOSITION, and nine low frequencies after decomposition is empty
Between surface electromyogram signal reconstruct, then extraction reconstruct after the multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned fuzzy entropy；
The multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned fuzzy entropy that step 3. obtains step 2 use principal component analysis
Method dimension-reduction treatment；
Step 3 is obtained myoelectricity characteristic vector and forms one group 12 with plantar pressure characteristic value and knee joint energy eigenvalue by step 4.
Dimensional feature vector；
Step 5. is walked the 12 dimensional feature vectors input particle group optimizing least square method supporting vector machine that step (4) is obtained
State mutually identifies；
Described particle group optimizing least square method supporting vector machine design is as follows：
The distribution characteristics of 12 dimensional feature vectors obtained according to step 4, least square branch is optimized by particle swarm optimization algorithm
Hold the kernel functional parameter and penalty factor of vector machine；Concretely comprise the following steps：
(1) iterations, initial position, speed and the current location of particle, the position of optimal particle in population are initialized；
(2) sample is divided into training set and test set, its fitness letter is determined according to least square method supporting vector machine categorised decision
Numerical value；
(3) fitness of each particle is calculated, p is updated according to fitness sizeiAnd pg, and more new particle is fast respectively according to following formula
Degree, particle position, inertia weight and Studying factors；
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<msubsup>
<mi>v</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mrow>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msubsup>
<mo>=</mo>
<mi>w</mi>
<msub>
<mi>v</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<msub>
<mi>r</mi>
<mn>1</mn>
</msub>
<mo>(</mo>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>-</mo>
<msubsup>
<mi>z</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>)</mo>
<mo>+</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
<msub>
<mi>r</mi>
<mn>2</mn>
</msub>
<mo>(</mo>
<msub>
<mi>p</mi>
<mrow>
<mi>g</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>-</mo>
<msubsup>
<mi>z</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>)</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<msubsup>
<mi>z</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mrow>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>z</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>v</mi>
<mrow>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mrow>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msubsup>
</mtd>
</mtr>
</mtable>
</mfenced>
Z in formulai=(zi1,zi2,...,zid), vi=(vi1,vi2,...,vid) position and the speed of i-th particle, d are represented respectively
For space dimensionality；K is iterations, and w is inertia weight, r1、r2For two random functions, c1、c2For scholar's factor；pi=
(pi1,pi2,...,pid) it is the optimal location that particle search arrives；pg=(pg1,pg2,...pgd) it is the optimal position that population searches
Put；
(4) if fitness meets desired output, optimizing terminates, otherwise by the particle rapidity after renewal, particle location information generation
Enter step 4- (3), until fitness meets desired output；The iterative algorithm of population optimizing more than, obtains making minimum
Two multiply the optimal value of the minimum penalty factor of SVMs error and kernel function radius parameter.
2. the lower limb gait phase identification method according to claim 1 based on multi-source bio signal, it is characterised in that step
1 detailed process is as follows：
Myoelectricity collection sensing is pasted respectively in right lateral thigh vastus medialis, rectus femoris, long adductor muscle and the tensor fasciae late muscle of subject
Device, electromyographic signal is obtained by myoelectricity acquisition system；It is special by obtaining plantar pressure installed in the plantar pressure sensor of sole
Value indicative；Acceleration transducer is placed respectively in the small leg outer side of thigh, when measuring lower limb difference posture from double-axel acceleration sensor
Acceleration axial component calculate knee joint angle characteristic value.
3. the lower limb gait phase identification method according to claim 1 based on multi-source bio signal, it is characterised in that：Step
3 detailed process is as follows：The multiple dimensioned energy value of surface electromyogram signal and multiple dimensioned fuzzy entropy totally 72 obtained from step (2)
Dimension, using principal component analytical method, according to the principle of gait phase character storage rate more than 95%, selection can be with from 72 dimensional features
Summarize the 10 dimensional features composition myoelectricity characteristic vector of primitive character value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711013763.XA CN107622260A (en) | 2017-10-26 | 2017-10-26 | Lower limb gait phase identification method based on multi-source bio signal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711013763.XA CN107622260A (en) | 2017-10-26 | 2017-10-26 | Lower limb gait phase identification method based on multi-source bio signal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107622260A true CN107622260A (en) | 2018-01-23 |
Family
ID=61093166
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711013763.XA Pending CN107622260A (en) | 2017-10-26 | 2017-10-26 | Lower limb gait phase identification method based on multi-source bio signal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107622260A (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108703824A (en) * | 2018-03-15 | 2018-10-26 | 哈工大机器人（合肥）国际创新研究院 | A kind of bionic hand control system and control method based on myoelectricity bracelet |
| CN109350071A (en) * | 2018-11-08 | 2019-02-19 | 华东师范大学 | A kind of flexible gait monitoring method calculated based on artificial intelligence |
| CN109645995A (en) * | 2019-01-16 | 2019-04-19 | 杭州电子科技大学 | Joint motions estimation method based on myoelectricity model and Unscented kalman filtering |
| CN109758274A (en) * | 2019-02-28 | 2019-05-17 | 清华大学 | Measurement method and system of the distal femur relative to prosthese gasket line of force track in single condyle knee replacements |
| CN109805928A (en) * | 2019-04-09 | 2019-05-28 | 太原理工大学 | A kind of cerebral apoplexy Method of EEG signals classification and system |
| EP3533392A1 (en) * | 2018-03-01 | 2019-09-04 | University of Eastern Finland | Electrode patch, system, and method for detecting indicator of parkinson's disease in person |
| CN110226932A (en) * | 2018-12-26 | 2019-09-13 | 杭州电子科技大学 | The plantar pressure feature extracting method of human body daily behavior movement |
| CN110236523A (en) * | 2019-06-17 | 2019-09-17 | 杭州电子科技大学 | Gait-Cardiac RR interval the correlating method returned based on Gauss |
| CN110427987A (en) * | 2019-07-17 | 2019-11-08 | 军事科学院系统工程研究院军需工程技术研究所 | A kind of the plantar pressure characteristic recognition method and system of arthritic |
| CN110537921A (en) * | 2019-08-28 | 2019-12-06 | 华南理工大学 | Portable gait multi-sensing data acquisition system |
| CN111053555A (en) * | 2019-12-31 | 2020-04-24 | 北京航空航天大学 | Terrain classification device and method based on surface electromyogram signals and plantar force |
| CN111476158A (en) * | 2020-04-07 | 2020-07-31 | 金陵科技学院 | Multi-channel physiological signal somatosensory gesture recognition method based on PSO-PCA-SVM |
| CN111513425A (en) * | 2020-06-04 | 2020-08-11 | 厉晨宇 | Detecting insole and walking posture adjusting system |
| CN111588376A (en) * | 2020-05-07 | 2020-08-28 | 北京联合大学 | Lower limb gait information extraction equipment based on electromyographic signals and angle signals |
| CN111643321A (en) * | 2020-04-30 | 2020-09-11 | 北京精密机电控制设备研究所 | Exoskeleton joint angle prediction method and system based on sEMG signals |
| CN111803250A (en) * | 2020-07-03 | 2020-10-23 | 北京联合大学 | Knee joint angle prediction method and system based on electromyographic signals and angle signals |
| CN112669568A (en) * | 2020-12-18 | 2021-04-16 | 浙江工商大学 | Multi-mode human body falling detection method |
| CN112754468A (en) * | 2021-01-07 | 2021-05-07 | 华南理工大学 | Human body lower limb movement detection and identification method based on multi-source signals |
| CN113520413A (en) * | 2021-08-25 | 2021-10-22 | 长春工业大学 | Lower limb multi-joint angle estimation method based on surface electromyogram signal |
| CN114947825A (en) * | 2022-04-29 | 2022-08-30 | 吉林大学 | Lower limb prosthesis continuous motion recognition method based on PSOGWO-SVM |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104107042A (en) * | 2014-07-10 | 2014-10-22 | 杭州电子科技大学 | Electromyographic signal gait recognition method based on particle swarm optimization and support vector machine |
| CN104537382A (en) * | 2015-01-12 | 2015-04-22 | 杭州电子科技大学 | Electromyographic signal gait recognition method for optimizing support vector machine based on genetic algorithm |
-
2017
- 2017-10-26 CN CN201711013763.XA patent/CN107622260A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104107042A (en) * | 2014-07-10 | 2014-10-22 | 杭州电子科技大学 | Electromyographic signal gait recognition method based on particle swarm optimization and support vector machine |
| CN104537382A (en) * | 2015-01-12 | 2015-04-22 | 杭州电子科技大学 | Electromyographic signal gait recognition method for optimizing support vector machine based on genetic algorithm |
Non-Patent Citations (2)
| Title |
|---|
| 刘磊: "基于多源信息的步态识别算法研究", 《中国优秀博士学位论文全文数据库信息科技辑》 * |
| 邹晓阳等: "基于多尺度模糊熵的动作表面肌电信号模式识别", 《生物医学工程学杂志》 * |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3533392A1 (en) * | 2018-03-01 | 2019-09-04 | University of Eastern Finland | Electrode patch, system, and method for detecting indicator of parkinson's disease in person |
| WO2019166557A1 (en) * | 2018-03-01 | 2019-09-06 | Itä-Suomen Yliopisto | Electrode patch, system, and method for detecting indicator of parkinson's disease in person |
| CN108703824A (en) * | 2018-03-15 | 2018-10-26 | 哈工大机器人（合肥）国际创新研究院 | A kind of bionic hand control system and control method based on myoelectricity bracelet |
| CN109350071A (en) * | 2018-11-08 | 2019-02-19 | 华东师范大学 | A kind of flexible gait monitoring method calculated based on artificial intelligence |
| CN110226932A (en) * | 2018-12-26 | 2019-09-13 | 杭州电子科技大学 | The plantar pressure feature extracting method of human body daily behavior movement |
| CN109645995A (en) * | 2019-01-16 | 2019-04-19 | 杭州电子科技大学 | Joint motions estimation method based on myoelectricity model and Unscented kalman filtering |
| CN109645995B (en) * | 2019-01-16 | 2021-09-07 | 杭州电子科技大学 | Joint motion estimation method based on electromyography model and unscented Kalman filtering |
| CN109758274B (en) * | 2019-02-28 | 2020-12-29 | 清华大学 | Method and system for measuring force line track of distal femur relative to prosthesis pad in unicondylar knee joint replacement |
| CN109758274A (en) * | 2019-02-28 | 2019-05-17 | 清华大学 | Measurement method and system of the distal femur relative to prosthese gasket line of force track in single condyle knee replacements |
| CN109805928A (en) * | 2019-04-09 | 2019-05-28 | 太原理工大学 | A kind of cerebral apoplexy Method of EEG signals classification and system |
| CN110236523A (en) * | 2019-06-17 | 2019-09-17 | 杭州电子科技大学 | Gait-Cardiac RR interval the correlating method returned based on Gauss |
| CN110427987A (en) * | 2019-07-17 | 2019-11-08 | 军事科学院系统工程研究院军需工程技术研究所 | A kind of the plantar pressure characteristic recognition method and system of arthritic |
| CN110537921A (en) * | 2019-08-28 | 2019-12-06 | 华南理工大学 | Portable gait multi-sensing data acquisition system |
| CN111053555B (en) * | 2019-12-31 | 2021-08-13 | 北京航空航天大学 | Terrain classification device and method based on surface electromyogram signals and plantar force |
| CN111053555A (en) * | 2019-12-31 | 2020-04-24 | 北京航空航天大学 | Terrain classification device and method based on surface electromyogram signals and plantar force |
| CN111476158A (en) * | 2020-04-07 | 2020-07-31 | 金陵科技学院 | Multi-channel physiological signal somatosensory gesture recognition method based on PSO-PCA-SVM |
| CN111476158B (en) * | 2020-04-07 | 2020-12-04 | 金陵科技学院 | Multi-channel physiological signal somatosensory gesture recognition method based on PSO-PCA-SVM |
| CN111643321A (en) * | 2020-04-30 | 2020-09-11 | 北京精密机电控制设备研究所 | Exoskeleton joint angle prediction method and system based on sEMG signals |
| CN111588376A (en) * | 2020-05-07 | 2020-08-28 | 北京联合大学 | Lower limb gait information extraction equipment based on electromyographic signals and angle signals |
| CN111513425A (en) * | 2020-06-04 | 2020-08-11 | 厉晨宇 | Detecting insole and walking posture adjusting system |
| CN111803250A (en) * | 2020-07-03 | 2020-10-23 | 北京联合大学 | Knee joint angle prediction method and system based on electromyographic signals and angle signals |
| CN112669568A (en) * | 2020-12-18 | 2021-04-16 | 浙江工商大学 | Multi-mode human body falling detection method |
| CN112754468A (en) * | 2021-01-07 | 2021-05-07 | 华南理工大学 | Human body lower limb movement detection and identification method based on multi-source signals |
| CN112754468B (en) * | 2021-01-07 | 2022-08-05 | 华南理工大学 | Human body lower limb movement detection and identification method based on multi-source signals |
| CN113520413A (en) * | 2021-08-25 | 2021-10-22 | 长春工业大学 | Lower limb multi-joint angle estimation method based on surface electromyogram signal |
| CN113520413B (en) * | 2021-08-25 | 2022-10-21 | 长春工业大学 | Lower limb multi-joint angle estimation method based on surface electromyogram signals |
| CN114947825A (en) * | 2022-04-29 | 2022-08-30 | 吉林大学 | Lower limb prosthesis continuous motion recognition method based on PSOGWO-SVM |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107622260A (en) | Lower limb gait phase identification method based on multi-source bio signal | |
| CN110537922B (en) | Human body walking process lower limb movement identification method and system based on deep learning | |
| Chen et al. | Locomotion mode classification using a wearable capacitive sensing system | |
| Huang et al. | An experimental powered lower limb prosthesis using proportional myoelectric control | |
| CN106308809A (en) | Method for recognizing gait of thigh amputation subject | |
| CN104107042A (en) | Electromyographic signal gait recognition method based on particle swarm optimization and support vector machine | |
| CN102722643A (en) | Multi-information merging movement mode identification method used for control of artificial lower surface electromyogram limb | |
| CN104656112B (en) | Based on surface electromyogram signal and the used personal localization method and devices combined of MEMS | |
| CN102764167B (en) | Myoelectric prosthesis control source lead optimization method based on correlation coefficients | |
| Look et al. | Dynamic stability of running: the effects of speed and leg amputations on the maximal Lyapunov exponent | |
| CN104983489A (en) | Road condition identifying method for lower limb prosthesis walking | |
| CN109446972A (en) | Gait recognition model establishing method, recognition method and device based on electromyographic signals | |
| CN109805935A (en) | A kind of intelligent waistband based on artificial intelligence hierarchical layered motion recognition method | |
| Huang et al. | Applications of machine learning techniques to a sensor-network-based prosthesis training system | |
| Wang et al. | Recognition of the Gait Phase Based on New Deep Learning Algorithm Using Multisensor Information Fusion. | |
| Yang et al. | Inertial sensing for lateral walking gait detection and application in lateral resistance exoskeleton | |
| Wang | Research on the improved CNN deep learning method for motion intention recognition of dynamic lower limb prosthesis | |
| Liu et al. | Deep learning based ground reaction force estimation for stair walking using kinematic data | |
| CN103720476A (en) | Stability evaluation method under curved path movement mode | |
| Negi et al. | Human locomotion classification for different terrains using machine learning techniques | |
| CN108325201A (en) | A kind of cadence detection method and equipment based on Multi-source Information Fusion | |
| CN107092865A (en) | A kind of new Gait Recognition system based on Kinect | |
| Chen et al. | An adaptive gait learning strategy for lower limb exoskeleton robot | |
| Lei et al. | Leg amputees motion pattern recognition based on principal component analysis and BP network | |
| Wu et al. | Gait recognition based on the plantar pressure data of ice and snow athletes |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180123 |
|
| RJ01 | Rejection of invention patent application after publication |