CN103886215B - Walking ability analyzing method and device based on muscle collaboration - Google Patents

Abstract

The invention provides a walking ability analyzing method and device based on muscle collaboration. The walking ability analyzing device based on muscle collaboration comprises a multi-channel surface myoelectricity acquisition module, a data analysis and storage module and a gait data analyzing and processing module. According to the walking ability calculating method based on muscle collaboration, the non-negative matrix factorization algorithm is adopted to extract gait muscle collaboration from a gait cycle surface myoelectricity signal, an adult gait muscle collaboration mode serves as a standard template, and the muscle collaboration mode of a subject is graded by means of a Synergy Comprehensive Calculation (SCC) template based on gait muscle collaboration mode similarity, so that characteristic parameters reflecting the gait of the subject are obtained. According to the walking ability analyzing method based on muscle collaboration, the difference of the walking abilities of different people is disclosed from the aspect of brain central nervous system control, and an objective walking ability parameter can be obtained. The objective walking ability parameter reflects the effect of control of the central nervous system over the gait.

Description

A kind of walking ability collaborative based on muscle analyzes method and apparatus

Technical field

The present invention is gait analysis technology based on sensor technology and bio signal treatment technology, refers in particular to a kind of based on flesh The walking ability analysis method and apparatus that meat is collaborative, belongs to sensor technology and bio signal processing technology field.

Background technology

Gait analysis has become one of analysis indispensable means of human motion system.Three-dimensional gait analysis technology at present There is two large problems: first gait analysis system uses dedicated video equipment and computer equipment, expensive, operates professional By force, it is difficult to universal.

Surface electromyogram signal (Surface Electromyography, SEMG) is obtained from body surface by surface electrode Electromyographic signal, gait data based on surface electromyogram signal obtains and need not special experimental site, can overcome gait analysis The shortcoming that video capture device is easily subject to environmental disturbances；And surface electromyogram signal can not only provide the essential information of gait, Can also extract muscle from surface electromyogram signal to work in coordination with, calculating for gait provides objective effective parameter.Therefore it provides it is a kind of Analyze method and apparatus based on the walking ability that muscle is collaborative to have great importance.

Summary of the invention

The deficiency existed for current gait analysis technology, it is contemplated that propose a kind of walking energy collaborative based on muscle Power analyzes method and apparatus.

A kind of walking ability collaborative based on muscle is analyzed method and is included:

(1) utilize Non-negative Matrix Factorization (Non-negative Matrix Factorization, NMF) algorithm from gait Period surface electromyographic signal (Surface Electromyography, SEMG) carries the gait muscle Cooperative Mode of step both legs. Refer specifically to NMF that (gait cycle Matrix of envelope is pretreatment and gait cycle to the gait cycle Matrix of envelope of experimenter's both legs Gait cycle surface electromyogram signal after segmentation) carry out Non-negative Matrix Factorization to extract left and right leg muscle Harmonious Matrix L, R.Will L, R are expressed as:

L=L^m×l={ l₁,l₂…l_l}

R=R^m×r={ r₁,r₂…r_r}

Wherein column vector l_i,r_j(i=1,2 ... l；J=1,2 ... r) representing the muscle Cooperative Mode of left and right lower limb respectively, l, r divide The gait muscle of left and right lower limb Biao Shi not work in coordination with quantity, m represents the quantity of one group of muscle selected in gait analysis；Use adult The muscle Cooperative Mode Criterion template set T that left and right lower limb is total:

T=T^m×s={ t₁,t₂…t_s}

Wherein column vector t_k(k=1,2 ... s) represent the flesh that standard gait muscle Cooperative Mode, i.e. adult's both legs are total Meat Cooperative Mode；S represents the muscle Cooperative Mode number that the quantity of standard gait muscle Cooperative Mode, i.e. adult left and right lower limb are total Amount；

(2) coordination with the synthesis based on gait muscle Cooperative Mode similarity is utilized to calculate (Synergy Comprehensive Calculation, SCC) calculating of model realization experimenter's walking ability index；SCC model considers experimenter's gait flesh Three factors of the symmetry that the quantity of meat Cooperative Mode, structure and left and right lower limb show in walking；And assist based on muscle Same walking ability analyzes method, and the angle controlled from nervous system discloses the brain coordination control strategy to gait motion, The result of calculation of this model can inherently reflect the brain coordination control ability to gait motion.

Described SCC model calculates walking ability and refers to calibration method, including step calculated below:

(1) by Pearson correlation coefficients computing formula R, (x y) calculates experimenter left and right each gait muscle Cooperative Mode of lower limb l_i,r_j(i=1,2 ... l；J=1,2 ... r) with each standard gait muscle Cooperative Mode t_k(k=1,2 ... coefficient R between s) (t_k,l_i),R(t_k,r_j).The similarity degree Sim (T, L), Sim of definition left and right leg muscle Harmonious Matrix L, R and standard form collection T (T, R), its computing formula is as follows:

$Sim(T,L)=\underset{i=1}{\overset{l}{\Σ}}F(T,l_i)$

$Sim(T,R)=\underset{j=1}{\overset{r}{\Σ}}F(T,r_j)$

F(T,l_i)=max (R (t_k,l_i)) k=1,2 ... s

F(T,r_j)=max (R (t_k,r_j)) k=1,2 ... s

F(T,l_i), F (T, r_j) represent left and right leg muscle Cooperative Mode l respectively_i,r_jSimilarity degree with standard form collection T. To each F (T, l_i), F (T, r_j) it is multiplied by score coefficient B=100/s, then left and right lower limb similarity score S_l,S_rIt is respectively as follows:

S_l=B*Sim (T, L)

S_r=B*Sim (T, R)

S_l,S_rValue between 0-100.

(2) can reflect that left and right lower limb is in gait due to the correlation coefficient between the lower limb gait muscle Cooperative Mode of experimenter left and right The symmetry shown in motion, therefore in symmetry score Sym defined in SCC model.Use Pearson's phase in the model (x y) calculates experimenter left and right lower limb each gait muscle Cooperative Mode l to close coefficient formulas R_i,r_j(i=1,2 ... l；J=1, 2 ... the coefficient R (r between r)_j,l_i) and define left and right leg muscle Harmonious Matrix L, the similarity degree Sim (R, L) between R, Its computing formula is:

$Sim(R,L)=\left\{\begin{array}{cc}\underset{i=1}{\overset{l}{\&Sigma;}}F(R,l_i)& r\&GreaterEqual;l\\ \underset{j=1}{\overset{r}{\&Sigma;}}F(L,r_j)& r<l\end{array}\right.$

F(R,l_i)=max (R (r_j,l_i)) j=1,2 ... r

F(L,r_j)=max (R (r_j,l_i)) i=1,2 ... l

F(R,l_i) represent left leg muscle Cooperative Mode l in this SCC model_iSimilarity degree with right leg muscle Harmonious Matrix R； F(L,r_j) then represent right leg muscle Cooperative Mode r_jSimilarity degree with left leg muscle Harmonious Matrix L.The meter of symmetry score Sym Calculation formula is as follows:

Sym=A*Sim (R, L)

Wherein (l, r), the value of Sym is between 0-100 for symmetry score coefficient A=100/min.

Finally, S is sought_l,S_r, the meansigma methods of Sym, obtain overall scores S:

S=(S_l+S_r+Sym)/3

S is the result of calculation of SCC model, and value is between 0-100.

A kind of device realizing the above-mentioned walking ability analysis method collaborative based on muscle, employing techniques below scheme: This device is made up of multi-channel surface myoelectric acquisition module, data parsing memory module, gait data analysis and processing module.Wherein Multi-channel surface myoelectric acquisition module is for gathering the SEMG of both legs in gait motion, and will believe by the way of wired or wireless Number it is sent to data parsing memory module；Data parsing memory module resolves gait data in real time, and stores；Gait number It is analyzed processing according to analysis and processing module gait data of load store from data parsing memory module, calculates walking ability Index.

Described gait analysis processing module include gait data pretreatment unit 1., gait cycle cutting unit 2., gait Muscle work in coordination with extraction unit 3., walking ability indicator calculating unit 4., display control unit 5.；The most 4. refer to as walking ability The core cell that mark calculates；Display control unit is 5. for arranging the running parameter of system, duty, and result Display etc..

The most 4. described walking ability index calculates is four serial computing unit, and wherein gait data pretreatment unit is 1. Process for the SEMG of every passage is filtered, goes average and rectification, to eliminate baseline noise in gait data acquisition etc. Impact, finally exports the envelope signal of SEMG.2. gait cycle cutting unit uses gait cycle based on acceleration signal to divide Cut algorithm and the envelope of SEMG after pretreatment is carried out the segmentation of gait cycle；Passage SEMG every to all gait cycles after segmentation Amplitude is normalized, poor to eliminate the signal amplitude between the different experimenters caused by factors such as electrode placement positions Different, finally by down-sampled for each passage of all gait cycle SEMG signals being partitioned into for uniform length, to be walked State cycle Matrix of envelope.Gait muscle is worked in coordination with extraction unit and 3. the gait cycle Matrix of envelope being partitioned into is carried out nonnegative matrix and divide Solve, calculate left and right lower limb gait muscle Cooperative Mode and the quantity of muscle Cooperative Mode thereof.Walking ability indicator calculating unit is the most then Use the calculating of SCC model realization walking ability index.

Described data parsing memory module and gait data analysis and processing module can be at common computer or band WIFI Smart mobile phone, the mobile terminal such as panel computer realizes, and therefore this walking ability analytical equipment is not limited by use occasion, can In hospital, rehabilitation institution, the environment such as family uses.

Accompanying drawing explanation

Fig. 1 is the structural representation of apparatus of the present invention；

Fig. 2 is surface myoelectric electrode setting method schematic diagram in the embodiment of the present invention；

Fig. 3 is the general frame calculating walking ability index in the present invention；

Fig. 4 is the schematic diagram of gait cycle signal segmentation in the embodiment of the present invention；

Fig. 5 is SCC model calculation flow chart in the embodiment of the present invention.

Detailed description of the invention

Illustrating below in conjunction with the accompanying drawings, providing one embodiment of the present of invention, concrete implementation process is as follows:

1., according to step 1 shown in Fig. 3, in this case study on implementation, we are to choose left and right lower limb tibialis anterior (TA), flatfish (LG), vastus lateralis (VL) outside flesh (SO), gastrocnemius, rectus femoris (RF), semitendinosus m. (SE), caput longum musculi bicipitis femoris (BF) and wealthy Tensor fascia femoris (TF) is as a example by totally 16 pieces of muscle are as the acquisition position of gait SEMG signal.Therefore the multi-channel surface shown in Fig. 1 Myoelectricity acquisition module (1) should include at least 16 channel surface electromyographic electrodes (each 8 passages of left and right lower limb).As in figure 2 it is shown, this embodiment Electrode used therein uses two strip electrode rods with skin contact, and during placement, two electrode connecting lines are consistent with muscle fiber direction of travel, Shave off fine hair at electrode placement positions in advance and use alcohol swab wiping.Use in this embodiment based on acceleration signal simultaneously The gait cycle segmentation of ACC, therefore multi-channel surface myoelectric acquisition module is also integrated with 2 passage 3-axis acceleration sensors, real It is individually positioned in below the lower limb knee of experimenter left and right at tibia during executing.Experimenter holds according to the speed of oneself the most comfortable Row walking task, in implementation process, experimenter need to move ahead in level ground N number of gait cycle, during acquisition module synchronous acquisition walking SEMG and ACC signal and send by the way of wired to the data parsing memory module (2) shown in Fig. 1.

2. the data parsing memory module (2) of device realizes on a common computer, first according to the concrete frame of data Form carries out real-time parsing, then parsed data is stored in file.

3. gait data analysis and processing module (3) realizes on a common computer.First its display control unit starts literary composition Part is loaded into function, loads pending gait data, has loaded when file has loaded interface prompt, and display control unit opens Dynamic data analysis, detailed process is as follows:

(1) original gait data initially enters pretreatment unit, carries out gait signal according to step 2 as shown in Figure 3 and locates in advance Reason: in this case study on implementation, first use cut-off frequency be 40Hz, the FIR high pass filter on the 50 rank gait data to experimenter Be filtered, then go average, rectification, finally to use cut-off frequency be 10Hz, the FIR low pass filter on 50 rank again filter with Obtain the gait SEMG envelope as shown in B in Fig. 4.

(2) gait SEMG envelope data enters gait cycle cutting unit, utilizes ACC signal to split it: such as C in Fig. 4 Shown in, in walking, acceleration signal ACC is quasi-periodic time varying signal, utilizes based on window function the most in this embodiment The peak point of peak-value detection method detection left and right lower limb acceleration transducer Z axis signal, is respectively mapped to left and right lower limb by peak point Gait SEMG envelope, is partitioned into N number of gait cycle envelope signal.B, C, D in Fig. 4 illustrates and utilizes gait ACC signal to split The detailed process of gait cycle.Finally by unified for every passage of N number of gait cycle envelope signal down-sampled for length n=500, and This passage is normalized by the maximum utilizing the every passage of gait cycle, obtains for extracting the gait week that muscle is collaborative Phase Matrix of envelope, as shown in the E in Fig. 4, the corresponding muscle passage of every a line of matrix.N number of step by lower limb single after pretreatment State cycle Matrix of envelope is designated as EMG_o={ O_i, i=1,2 ... N}, O_iBeing the matrix of a m × n, m represents selected in gait analysis The quantity of the one group of muscle taken, m=8 in this embodiment.

(3) gait muscle works in coordination with extraction unit to gait cycle Matrix of envelope EMG_oProcess, the step 3 of Fig. 3 know, Gait muscle works in coordination with extraction unit can extract the muscle Cooperative Mode of left and right lower limb respectively, and calculates the quantity of Cooperative Mode, and it is detailed Thin process is as follows:

1) to each gait cycle matrix O_i, according to Non-negative Matrix Factorization NMF algorithm, calculating muscle and working in coordination with quantity is j (j=1,2 ... m) corresponding muscle Harmonious Matrix W_i ^m×j,

2) gait muscle Cooperative Mode quantity s is sought: utilize W_i ^m×j,Trying to achieve muscle Cooperative Mode quantity is reconstruction during j Gait pattern matrix stackWherein rebuild gait pattern matrixThen institute is calculated There is EMG_jWith EMG before decomposition_oSquare error matrix stack V={V_j, j=1,2 ... m}, wherein V_jComputing formula is:

$\begin{array}{cc}{V}_j=\frac{1}{N}\underset{i=1}{\overset{N}{\&Sigma;}}(1-\frac{{(O_i-R_i^j)}^2}{{O_i}^2})& 0<V_j<1\end{array}$

V_jRepresent that muscle works in coordination with reconstruction gait pattern matrix stack EMG when quantity is j_jWith original gait cycle matrix stack EMG_o Mean Square Error.According to priori, gait muscle Cooperative Mode quantity s is defined as V under t inspection (p < 0.05)_j's Value works in coordination with quantity j more than the minimum muscle corresponding to 0.95, and the muscle Harmonious Matrix of experimenter is:

$W^{m\&times;s}=\{w_1,w_2...w_s\}=\frac{1}{N}\underset{i=1}{\overset{N}{\&Sigma;}}{W}_i^{m\&times;s}$

3) experimenter left and right leg muscle Harmonious Matrix is expressed as:

L=L^m×l={ l₁,l₂…l_l}

R=R^m×r={ r₁,r₂…r_r}

Wherein column vector l_i,r_j(i=1,2 ... l；J=1,2 ... r) representing the muscle Cooperative Mode of left and right lower limb respectively, l, r divide The gait muscle of left and right lower limb Biao Shi not work in coordination with quantity.Extract in advance about several adult according to method 1-3 described in this embodiment The muscle Cooperative Mode Criterion template set T that lower limb is total:

T=T^m×s={ t₁,t₂…t_s}

Wherein column vector t_k(k=1,2 ... s) represent standard gait muscle Cooperative Mode, s represents that standard gait muscle is worked in coordination with The quantity of pattern, and standard form is stored in Data Analysis Services module.

(4) walking ability indicator calculating unit is according to SCC model, calculates walking ability according to flow chart as shown in Figure 5: First gait muscle is worked in coordination with result of calculation L of extraction unit, R, l, r and standard form collection T and standard gait muscle works in coordination with mould Formula quantity s inputs this unit；Calculate left and right lower limb respective similarity score S the most respectively_l,S_r；According to the comparative result of l, r, meter Calculate left and right lower limb symmetry score Sym in walking；Last comprehensive S_l,S_r, Sym obtains walking ability index.Its concrete calculating Process is as follows:

1) left and right lower limb similarity score S is calculated_l,S_rShi Xianyong Pearson correlation coefficients computing formula R (x, y) calculate tested Person left and right lower limb each gait muscle Cooperative Mode l_i,r_j(i=1,2 ... l；J=1,2 ... r) work in coordination with mould with each standard gait muscle Formula t_k(k=1,2 ... coefficient R (t between s)_k,l_i),R(t_k,r_j), try to achieve l by equation below_i,r_jWith standard form collection T Similarity degree F (T, l_i), F (T, r_j):

F(T,l_i)=max (R (t_k,l_i)) k=1,2 ... s

F(T,r_j)=max (R (t_k,r_j)) k=1,2 ... s

$R(x,y)=\frac{\underset{i=1}{\overset{size\left(x\right)}{\&Sigma;}}(x_i-\stackrel{\&OverBar;}{x})(y_i-\stackrel{\&OverBar;}{y})}{\sqrt{\underset{i=1}{\overset{size\left(x\right)}{\&Sigma;}}{(x_i-\stackrel{\&OverBar;}{x})}^2}\sqrt{\underset{i=1}{\overset{size\left(x\right)}{\&Sigma;}}{(y_i-\stackrel{\&OverBar;}{y})}^2}}$

Then to each F (T, l_i), F (T, r_j) being multiplied by score coefficient B=100/s, then left and right lower limb is similar to standard form Property score S_l,S_rIt is respectively as follows:

$S_l=B\underset{i=1}{\overset{l}{\&Sigma;}}F(T,l_i)$

$S_r=B\underset{j=1}{\overset{r}{\&Sigma;}}F(T,r_j)$

WhereinRepresent the similarity degree Sim (T, L), Sim of L, R and T in SCC model respectively (T,R)；S_l,S_rValue between 0-100.

2), when calculating left and right lower limb symmetry score, first by Pearson correlation coefficients computing formula R, (x y) calculates experimenter left Right lower limb each gait muscle Cooperative Mode l_i,r_j(i=1,2 ... l；J=1,2 ... the coefficient R (r between r)_j,l_i), when r >= L, left and right leg muscle Harmonious Matrix L, the similarity degree Sim (R, L) between R be:

$Sim(R,L)=\underset{i=1}{\overset{l}{\&Sigma;}}F(R,l_i)$

F(R,l_i)=max (R (r_j,l_i)) j=1,2 ... r

When r < during l, left and right leg muscle Harmonious Matrix L, the similarity degree Sim (R, L) between R be:

$Sim(R,L)=\underset{j=1}{\overset{l}{\&Sigma;}}F(L,r_j)$

F(L,r_j)=max (R (r_j,l_i)) i=1,2 ... j

Sim (R, L) and symmetry score coefficient A=100/min (l, r) is multiplied, then symmetry score Sym is:

Sym=A*Sim (R, L)

Last walking ability indicator calculating unit calculates S_l,S_r, the meansigma methods of Sym, obtain overall scores S:

S=(S_l+S_r+Sym)/3

The walking ability index S calculated by the method can reflect the difference of walking ability between Different Individual.

Claims (9)

1. analyze method based on the walking ability that muscle is collaborative for one kind, it is characterised in that this walking ability analyze method first with Non-negative Matrix Factorization (Non-negative Matrix Factorization, NMF) algorithm is from gait cycle surface electromyogram signal (Surface Electromyography, SEMG) extracts the gait muscle Cooperative Mode of both legs；Then utilize based on gait The coordination with the synthesis of muscle Cooperative Mode similarity calculates (Synergy Comprehensive Calculation, SCC) model meter Calculate the walking ability of Different Individual；Wherein SCC model is using the stable gait muscle Cooperative Mode of adult as standard gait Muscle Cooperative Mode, the comprehensive similarity analyzing experimenter left and right leg muscle Cooperative Mode and standard gait muscle Cooperative Mode, And the symmetry that left and right lower limb shows in gait motion, calculate the characteristic parameter of reflection walking ability；

Described walking ability is analyzed method and is specifically included following steps:

(1) utilize Non-negative Matrix Factorization NMF algorithm from gait cycle surface electromyogram signal (Surface Electromyography, SEMG) extracting the gait muscle Cooperative Mode of both legs, the left and right leg muscle Harmonious Matrix L that will extract in, R is expressed as:

L=L^m×l={ l₁,l₂…l_l}

R=R^m×r={ r₁,r₂…r_r}

Wherein column vector l_i,r_jThe muscle Cooperative Mode of expression left and right lower limb respectively, described i=1,2 ... l；J=1,2 ... r, l, r divide The gait muscle of left and right lower limb Biao Shi not work in coordination with quantity, m represents the quantity of one group of muscle selected in gait analysis；By adult The muscle Cooperative Mode that both legs have is as standard form collection T:

T=T^m×s={ t₁,t₂…t_s}

T is the matrix of a m × s, column vector t_kExpression standard gait muscle Cooperative Mode, described k=1,2 ... s, s represent standard The quantity of gait muscle Cooperative Mode；

(2) coordination with the synthesis based on gait muscle Cooperative Mode similarity is utilized to calculate (Synergy Comprehensive Calculation, SCC) model, calculating walking ability index S:

1) by Pearson correlation coefficients computing formula R, (x y) calculates experimenter left and right lower limb each gait muscle Cooperative Mode l_i,r_j With each standard gait muscle Cooperative Mode t_kBetween coefficient R (t_k,l_i),R(t_k,r_j)；Definition left and right leg muscle works in coordination with square Similarity degree Sim (T, L), the Sim (T, R) of battle array L, R and standard form collection T, its computing formula is as follows:

$Sim(T,L)=\underset{i=1}{\overset{l}{\&Sigma;}}F(T,l_i)$

$Sim(T,R)=\underset{j=1}{\overset{r}{\&Sigma;}}F(T,r_j)$

F(T,l_i)=max (R (t_k,l_i)) k=1,2 ... s

F(T,r_j)=max (R (t_k,r_j)) k=1,2 ... s

F(T,l_i), F (T, r_j) represent left and right leg muscle Cooperative Mode l respectively_i,r_jWith the similarity degree of standard form collection T, to often Individual F (T, l_i), F (T, r_j) it is multiplied by score coefficient B=100/s, then left and right lower limb similarity score S_l,S_rIt is respectively as follows:

S_l=B*Sim (T, L)

S_r=B*Sim (T, R)

S_l,S_rValue between 0-100；

2) in SCC model, by Pearson correlation coefficients computing formula R, (x y) calculates experimenter left and right each gait muscle of lower limb association With pattern l_i,r_jBetween coefficient R (r_j,l_i) and define left and right leg muscle Harmonious Matrix L, the similarity degree Sim between R (R, L), its computing formula is:

$Sim(R,L)=\left\{\begin{array}{cc}\underset{i=1}{\overset{l}{\&Sigma;}}F(R,l_i)& r\&GreaterEqual;l\\ \underset{j=1}{\overset{r}{\&Sigma;}}F(L,r_j)& r<l\end{array}\right.$

F(R,l_i)=max (R (r_j,l_i)) j=1,2 ... r

F(L,r_j)=max (R (r_j,l_i)) i=1,2 ... l

F(R,l_i) represent left leg muscle Cooperative Mode l in this SCC model_iSimilarity degree with right leg muscle Harmonious Matrix R；F(L, r_j) then represent right leg muscle Cooperative Mode r_jWith the similarity degree of left leg muscle Harmonious Matrix L, the calculating of symmetry score Sym is public Formula is as follows:

Sym=A*Sim (R, L)

(l, r), the value of Sym is between 0-100 for the coefficient A=100/min of wherein said symmetry score；

3) last, seek S_l,S_r, the meansigma methods of Sym, obtain overall scores S:

S=(S_l+S_r+Sym)/3

S is the result of calculation of SCC model, and the value of S is between 0-100.

2., for performing a walking ability analytical equipment collaborative based on muscle for method described in claim 1, its feature exists In, described analytical equipment includes multi-channel surface myoelectric acquisition module (1), data parsing memory module (2), gait data analysis Processing module (3)；Wherein multi-channel surface myoelectric acquisition module (1) communicates with data parsing memory module (2)；Gait data divides The gait data of storage in data parsing memory module (2) is analyzed processing by analysis processing module (3), calculates walking ability and refers to Mark.

3. walking ability analytical equipment as claimed in claim 2, it is characterised in that described multi-channel surface myoelectric acquisition module (1) for gathering the surface electromyogram signal of both legs in gait, owing to gait motion relates to polylith lower limb muscles, module should at least be wrapped Surface myoelectric electrode containing 16 passages；The data of multi-channel surface myoelectric acquisition module are sent to by the way of wired or wireless Data parsing memory module (2)；Data parsing memory module (2) completes the parsing storage of gait data in real time.

4. walking ability analytical equipment as claimed in claim 2, it is characterised in that described gait data analysis and processing module (3) include four serial computing unit: gait data pretreatment unit 1., gait cycle cutting unit 2., gait muscle works in coordination with Extraction unit 3., walking ability indicator calculating unit 4.；The most 4. the core cell calculated as walking ability index；Additionally Including display control unit 5., it is for arranging the running parameter of system, duty, and the display of result.

5. walking ability analytical equipment as claimed in claim 4, it is characterised in that gait data pretreatment unit is 1. from data Resolve and memory module (2) load original gait SEMG signal, and complete the smooth of every passage SEMG signal and filter preprocessing, To eliminate the impact of baseline noise in gait data acquisition, output is the envelope signal of SEMG.

6. walking ability analytical equipment as claimed in claim 4, it is characterised in that gait cycle cutting unit 2. use based on The gait cycle partitioning algorithm of acceleration signal carries out the segmentation of gait cycle to the envelope signal of SEMG after pretreatment；After segmentation The amplitude of passage SEMG every to all gait cycles is normalized, and causes not by electrode placement positions factor to eliminate With the signal amplitude difference between experimenter；Finally by down-sampled for all gait cycle signals being partitioned into for uniform length, defeated Go out gait cycle Matrix of envelope.

7. walking ability analytical equipment as claimed in claim 4, it is characterised in that gait muscle is worked in coordination with extraction unit and 3. used Both legs gait cycle Matrix of envelope is decomposed by Non-negative Matrix Factorization NMF algorithm, to extract both legs gait muscle Cooperative Mode, And export the muscle Harmonious Matrix L of left and right lower limb, Cooperative Mode quantity l of R and correspondence, r；It it is critical only that and utilizes NMF algorithm to carry The Harmonious Matrix taken rebuilds gait pattern, meets constraints when rebuilding the gait pattern error amount with original gait, then tries to achieve Quantity that gait muscle is collaborative and gait muscle Cooperative Mode.

8. walking ability analytical equipment as claimed in claim 4, it is characterised in that walking ability indicator calculating unit is first Load standard form collection T, then with SCC computation model, walking ability index is calculated, export result of calculation.

9. walking ability analytical equipment as claimed in claim 2, it is characterised in that described data parsing memory module (2) and Gait data analysis and processing module (3), at common computer or the smart mobile phone of band WIFI, panel computer realizes.