CN105631195A - Wearable multi-information fusion gait analysis system and method thereof - Google Patents

Abstract

The invention discloses a wearable multi-information fusion gait analysis system and a method thereof. The wearable multi-information fusion gait analysis system is characterized by comprising a flexible sensor module, a three-axis acceleration and gyro sensor, a microprocessor, a power supply module, a transmission module, a positioning module, a client side, a cloud server and the like. The flexible sensor is used for measuring the gait data of the bottoms of feet when a user walks, the gait data is sent to the cloud server after being analyzed through the microprocessor, and a gait situation is displayed on a user client side in real time. If the user is under an emergency situation, the guardian of the user can receive alarming information and position the position of the user in real time so as to rescue the user. The wearable multi-information fusion gait analysis system can cause the guardian to know the gait situation of a person under guardianship at any time, and receives the alarming information and the positioning information generated when the person under guardianship moves, so that support and help can be provided for the rehabilitation of a foot patient, the gait monitoring of the aged, walk learning of children and the like.

Description

The gait analysis system of a kind of wearable Multi-information acquisition and method thereof

Technical field

The present invention relates to intelligence wearable device technical field, particularly relate to gait analysis system and the method thereof of a kind of wearable Multi-information acquisition.

Background technology

Along with growth in the living standard, the health condition of self is increasingly paid attention to by people, particularly sport health aspect, and people also increasingly wish that the moment understands oneself moving situation. Along with the rise of smart mobile phone, the development of intelligence wearable technology and big data technique, the demand of people is also achieved.

Gait analysis can apply in sport health scientific domain. Such as: extracted by the gait feature parameter that the normal walking posture of old people is worsened and propose that corresponding preventive measure can efficiently reduce old people fall down number of times; By the gait feature of hemiplegic patient has been studied, such that it is able to effectively analyze the recovery that hemiplegic patient is current; By the gait analysis to landform such as stair, level road, landslide, steps, and then effectively can provide more theoretical foundation for the research of artificial intelligence's lower limb; Additionally, gait analysis patient that can also be used to suffering from some disease in health science field is identified, such as Parkinsonian, patients with cerebral palsy, diabetics etc.

At present, both at home and abroad a lot of research be there has also been for gait analysis. TaoLiu et al. places gyroscope and accelerometer on the foot of gauger, shank bottom and knee, thus measuring the state of lower limb walking. KevinDeschamps et al. is foot bottom force-bearing situation when walking measured by sole placement force sensor. These study the direction provided for gait analysis, and proposition can finer measuring method.

Summary of the invention

The present invention is the weak point overcoming prior art to exist, propose a kind of can the gait analysis system of wearable Multi-information acquisition of man-machine interaction and method thereof, the present invention can allow guardian understand the gait situation of custodial person at any time, real-time reception custodial person motion in warning message and location information, thus being foot patients ' recovery, old people walks appearance monitoring, and toddler etc. provides supports and help.

The present invention adopts the following technical scheme that to achieve the above object of the invention

The feature of the gait analysis system of a kind of wearable Multi-information acquisition of the present invention includes: flexible sensor module, 3-axis acceleration and gyro sensor, microprocessor, supply module, transport module, locating module, client and Cloud Server; It is made up of gait analysis device described 3-axis acceleration and gyro sensor, microprocessor, supply module and transport module and is arranged on the arch place of sole; Described client includes guardian's client and subscription client;

Described sensor assembly includes pliable pressure sensor and soft stretch sensor;

Described pliable pressure sensor is arranged on the heel portion of shoe pad, arch portion and forward foot in a step metacarpus, is used for gathering foot force information and passing to described microprocessor by adsorption-type pricking with needle;

Described soft stretch sensor is arranged between the arch portion of shoe pad and forward foot in a step metacarpus, is used for gathering plantar flex degree and passing to described microprocessor also by adsorption-type pricking with needle;

Described 3-axis acceleration and gyro sensor are for gathering acceleration on three directions and angle of inclination and passing to described microprocessor;

Described locating module obtains the positional information of user and is sent to described microprocessor;

Received foot force information is calculated by described microprocessor, it is thus achieved that Center of Pressure dot information; And be analyzed according to described Center of Pressure dot information, plantar flex degree and acceleration, it is thus achieved that gait feature value;

Described gait feature value is input in neutral net and carries out gait judgement by described microprocessor, it is thus achieved that gait state result be sent to described client by described transport module, or be sent to described Cloud Server by described transport module and store; Make described client can obtain gait state result from described Cloud Server;

Described microprocessor is sent to described Cloud Server by described transport module after described positional information is processed and stores;

Described angle of inclination is analyzed by described microprocessor, it may be judged whether fall, if falling, then sends warning message to described Cloud Server by transport module; Described alarm information pushing is given described guardian's client by described Cloud Server again;

Described guardian's client obtains customer position information by described Cloud Server.

The feature of the gait analysis system of wearable Multi-information acquisition of the present invention lies also in,

The composition of described pliable pressure sensor includes: flexible PCB, copper electrode, micro structure sensitive material and plastic sheeting;

Flexible PCB is printed with described copper electrode on described, and covers described micro structure sensitive material on the surface of described copper electrode; Described micro structure sensitive material covers described plastic sheeting;

The composition of described soft stretch sensor includes: overlying plastic thin film, conductive silver glue, sensitive material and lower floor's plastic sheeting;

Described lower floor plastic sheeting is printed with spaced two pieces of conductive silver glues as electrode; Described sensitive material is pasted onto on described lower floor plastic sheeting by described two pieces of conductive silver glues; Overlying plastic thin film is covered on the surface of described sensitive material.

Described micro structure sensitive material is set to pyramid array format, and contacts with copper electrode with tower top.

Described micro structure sensitive material is after being mixed by rare to white carbon black and graphite mass ratio with 3:1, then is filled in silicone rubber molding with the total mass fraction of 4% and obtains;

Described sensitive material is after being mixed by rare to white carbon black and graphite mass ratio with 3:1, then is filled in silicone rubber molding with the total mass fraction of 6% and obtains.

The feature of the gait analysis method of a kind of wearable Multi-information acquisition of the present invention is to carry out as follows:

Step 1, it is set to initial point O with a summit of the boundary rectangle of shoe pad, two adjacent sides of described initial point O is respectively set to X-axis and Y-axis, using the direction of vertical and described boundary rectangle as Z axis, constitute coordinate system O-XYZ;

Step 2, in described coordinate system XOY, obtain the position coordinates of n pliable pressure sensor, be designated as { (x₁,y₁),(x₂,y₂),��,(x_i,y_i),��,(x_n,y_n), (x_i,y_i) represent i-th pliable pressure sensor position coordinates; 1��i��n;

Step 3, utilize described n pliable pressure sensor to obtain n foot force value, be designated as { P₁,P₂,��,P_i,��,P_n; P_iRepresent the foot force value of i-th pliable pressure sensor;

Step 4, formula (1) and formula (2) is utilized to obtain Center of Pressure dot information (x_c,y_c):

$x_c=\underset{i=1}{\overset{n}{\Σ}}(x_i\×P_i)/\underset{i=1}{\overset{n}{\Σ}}{P}_i---\left(1\right)$

$y_c=\underset{i=1}{\overset{n}{\Σ}}(y_i\×P_i)/\underset{i=1}{\overset{n}{\Σ}}{P}_i---\left(2\right)$

Step 5, by soft stretch sensor (2) obtain t plantar flex degree C_t;

Step 6, obtain the acceleration on three directions of t by 3-axis acceleration and gyro sensorAnd utilize formula (3) to obtain the vector value S of t acceleration_t:

$S_t=\sqrt{{\left(a_x^t\right)}^2+{\left(a_y^t\right)}^2+{\left(a_z^t\right)}^2}---\left(3\right)$

When step 7, respectively collection normal gait and abnormal gait, Center of Pressure dot information (x_c,y_c), plantar flex degree C_t, acceleration on three directionsAnd the vector value S of acceleration_t, the respective seven kinds of eigenvalues of these seven kinds of data, for being trained neutral net, it is thus achieved that gait analysis model; Seven kinds of eigenvalues include maximum, minima, average, excursion, amplitude, variance and standard deviation;

Step 8, set described Center of Pressure dot information (x_c,y_c), plantar flex degree, acceleration on three directions and acceleration the respective seven kinds of eigenvalues of vector value walk accordingly threshold value;

Step 9, with T for the sampling period, F is sample frequency, formed sampling time window; In sampling time window, gather by Center of Pressure dot information (x_c,y_c), plantar flex degree, acceleration on three directions and acceleration the exercise data that forms of the respective seven kinds of eigenvalues of vector value;

Step 10, described exercise data and set walking threshold value are compared, it is judged that in sampling time window, whether user is at walking states; If at walking states, then perform step 11;

Step 11, described exercise data is inputted in described gait analysis model, thus obtaining the gait state result in sampling time window.

Compared with the prior art, beneficial effects of the present invention is embodied in:

1, many heat transfer agents are merged by the present invention, and support radio communication, when being walked by flexible sensor measurement user, the gait data of foot bottom, after microprocessor analysis, sent to Cloud Server by transport module, and show gait situation at subscription client in real time, if user is in an emergency, user guardian also can receive warning message real-time positioning customer location, implement rescue, thus being foot patients ' recovery, old people walks appearance monitoring, and toddler etc. provides a kind of worn for long periods monitoring equipment.

2, the pliable pressure sensor of the present invention, is do substrate with flexible PCB, adopts interdigitated electrode structure, using the rare mixing filled silicon rubber of white carbon black and graphite as sensing unit, and adopt small pyramid structure at cell surface. Compared to conventional pressure sensor, the feature such as this pliable pressure sensing arrangement has Grazing condition, thickness is thin, pressure-sensitive character is strong, range is big, good stability. Can being comfortably positioned on shoe pad, the wearing not affecting wearer is experienced; And less costly, easily change.

3, the soft stretch sensor of the present invention, is with plastic sheeting for substrate, and conductive silver glue is as electrode, using the rare mixing filled silicon rubber of white carbon black and graphite as sensing unit. Compared to tradition stretch sensor. This soft stretch sensor has the feature such as Grazing condition, thin, the good stability of thickness.

4, the gait analysis method of the Multi-information acquisition of the present invention, adopt time slip-window method that signal is carried out Division Sampling, utilize Center of Pressure point method by the information fusion of multiple pressure transducers, in conjunction with stretch sensor and gyroscope, by the seven of three kinds of sensors kinds of feature fusion, achieve the convergence analysis to sole gait data, decrease amount of calculation.

Accompanying drawing explanation

Fig. 1 is pliable pressure sensor plane structure explanation figure of the present invention;

Fig. 2 is pliable pressure sensor up-down structure explanation figure of the present invention;

Fig. 3 is soft stretch sensor plane structure explanation figure of the present invention;

Fig. 4 is soft stretch sensor up-down structure explanation figure of the present invention;

Fig. 5 is multi information measuring and analysis system explanation figure of the present invention.

Fig. 6 is present system block diagram;

Number in the figure: 1 pliable pressure sensor; 1a flexible PCB, 1b copper electrode, 1c micro structure sensitive material; 1d plastic sheeting; 2 soft stretch sensors; 2a overlying plastic thin film; 2b conductive silver glue; 2c sensitive material; 2d lower floor plastic sheeting; 3 adsorption-type contact pins; 4 charging inlets.

Detailed description of the invention

In the present embodiment, as shown in Figure 6, the gait analysis system of a kind of wearable Multi-information acquisition includes: sensor assembly, 3-axis acceleration and gyro sensor, microprocessor, supply module, transport module, locating module, client and Cloud Server; It is made up of gait analysis device 3-axis acceleration and gyro sensor, microprocessor, supply module and transport module and is arranged on the arch place of sole; Client includes guardian's client and subscription client.

Sensor assembly includes pliable pressure sensor 1 and soft stretch sensor 2; Pliable pressure sensor and soft stretch sensor are all distributed on shoe pad, and the distributing position of pliable pressure sensor and soft stretch sensor is as shown in Figure 5.

Pliable pressure sensor 1 is arranged on the heel portion of shoe pad, arch portion and forward foot in a step metacarpus, is used for gathering foot force information and passing to microprocessor by adsorption-type pricking with needle 3; Pliable pressure sensor construction figure is as it is shown in figure 1, using flexible PCB 1a as substrate, copper electrode 1b adopts interdigital structure; Sensitive material 1c is after being mixed by rare to white carbon black and graphite mass ratio with 3:1, then is filled in silicone rubber with the total mass fraction of 4%; Pour the sensitive material after filling in particular mold sizing, make sensitive material bottom surface have pyramid shape microwave structure, and contact with copper electrode 1b with tower top. Pliable pressure sensor construction is from top to bottom as in figure 2 it is shown, be followed successively by: plastic sheeting 1d, sensitive material 1c, copper electrode 1b and flexible PCB 1a.

Soft stretch sensor 2 is arranged between the arch portion of shoe pad and forward foot in a step metacarpus, is used for gathering plantar flex degree and passing to microprocessor also by adsorption-type pricking with needle 3; Soft stretch sensor construction such as Fig. 3, adopts conductive silver glue to make electrode, and sensitive material 2c is after being mixed by rare to white carbon black and graphite mass ratio with 3:1, then with the total mass fraction of 6% acquisition that is filled in silicone rubber molding. Soft stretch sensor construction from top to bottom as shown in Figure 4, is followed successively by: overlying plastic thin film 2a, sensitive material 2c, conductive silver glue 2b and lower floor plastic sheeting 2d.

Lithium battery and charging module, 3-axis acceleration sensor and gyroscope, locating module, bluetooth module, GPRS module, as primary module, is positioned over arch bottom. Sensor assembly and primary module are connected by adsorption-type pricking with needle 3 interface.

3-axis acceleration and gyro sensor are for gathering acceleration on three directions and angle of inclination and passing to microprocessor;

Locating module obtains the positional information of user and is sent to microprocessor;

Received foot force information is calculated by microprocessor, it is thus achieved that Center of Pressure dot information; And be analyzed according to Center of Pressure dot information, plantar flex degree and acceleration, it is thus achieved that gait feature value, specifically, is carry out as follows:

Step 1, it is set to initial point O with a summit of the boundary rectangle of shoe pad, the two of initial point O adjacent sides is respectively set to X-axis and Y-axis, using the vertical direction with boundary rectangle as Z axis, composition coordinate system O-XYZ;

Step 2, in coordinate system XOY, obtain the position coordinates of n pliable pressure sensor, be designated as { (x₁,y₁),(x₂,y₂),��,(x_i,y_i),��,(x_n,y_n), (x_i,y_i) represent i-th pliable pressure sensor position coordinates; 1��i��n;

Step 3, utilize n pliable pressure sensor to obtain n foot force value, be designated as { P₁,P₂,��,P_i,��,P_n; P_iRepresent the foot force value of i-th pliable pressure sensor;

Step 4, formula (1) and formula (2) is utilized to obtain Center of Pressure dot information (x_c,y_c):

$x_c=\underset{i=1}{\overset{n}{\Σ}}(x_i\×P_i)/\underset{i=1}{\overset{n}{\Σ}}{P}_i---\left(1\right)$

$y_c=\underset{i=1}{\overset{n}{\Σ}}(y_i\×P_i)/\underset{i=1}{\overset{n}{\Σ}}{P}_i---\left(2\right)$

Step 5, obtained the plantar flex degree C of t by soft stretch sensor 2_t;

Step 6, obtain the acceleration on three directions of t by 3-axis acceleration and gyro sensorAnd utilize formula (3) to obtain the vector value S of t acceleration_t:

$S_t=\sqrt{{\left(a_x^t\right)}^2+{\left(a_y^t\right)}^2+{\left(a_z^t\right)}^2}---\left(3\right)$

When step 7, respectively collection normal gait and abnormal gait, Center of Pressure dot information (x_c,y_c), plantar flex degree C_t, acceleration on three directionsAnd the vector value S of acceleration_t, the respective seven kinds of eigenvalues of these seven kinds of data, for being trained neutral net, it is thus achieved that gait analysis model; Seven kinds of eigenvalues include maximum, minima, average, excursion, amplitude, variance and standard deviation;

These seven kinds of eigenvalues represent respectively:

Maximum: the maximum in all sampled points;

Minima: the maximum in all sampled points;

Average: the meansigma methods of all sampled points;

Excursion: the difference of maxima and minima in all sampled points;

Amplitude: the difference of maximum and average in all sampled points;

Variance: the variance of all sampled points;

Standard deviation: the standard deviation of all sampled points.

Step 8, setting pressure central point information (x_c,y_c), plantar flex degree C_t, acceleration on three directionsAnd the vector value S of acceleration_tRespective seven kinds of eigenvalues are walked threshold value accordingly;

Step 9, with T for the sampling period, F is sample frequency, formed sampling time window; In sampling time window, gather by Center of Pressure dot information (x_c,y_c), plantar flex degree C_t, acceleration on three directionsAnd the vector value S of acceleration_tThe exercise data that respective seven kinds of eigenvalues form;

Step 10, exercise data and set walking threshold value are compared, it is judged that in sampling time window, whether user is at walking states; If at walking states, then perform step 11;

Step 11, exercise data is inputted in gait analysis model, thus obtaining the gait state result in sampling time window.

Gait feature value is input in neutral net and carries out gait judgement by microprocessor. The three layers BP network built is: ground floor is input layer, being made up of 49 input nodes, it is the maximum of seven kinds of respective seven kinds of eigenvalues of data, minima, average, excursion, amplitude, variance, standard deviation that 49 corresponding respectively input components are corresponding in turn to. 49 input components constitute 1 input vector X and are:; The second layer is hidden layer; Third layer is output layer, it is made up of 2 output nodes, two output nodes 2 corresponding respectively output components represent that gait is correct and abnormal gait successively, 2 output components constitute 1 output vector Y and are:, to be characterized as gait correct status, to be characterized as abnormal gait state.

System block diagram is as shown in Figure 6, it is thus achieved that gait state result be sent to client by transport module, or be sent to Cloud Server by transport module and store; Make client can obtain gait state result from Cloud Server;

Microprocessor is sent to Cloud Server by transport module after positional information is processed and stores;

Angle of inclination is analyzed by microprocessor, it may be judged whether fall, if falling, then sends warning message to Cloud Server by transport module; Cloud Server again by alarm information pushing to guardian's client;

Guardian's client obtains customer position information by Cloud Server.

Mechanical periodicity in being embodied as, according to foot's acceleration and foot force, it is possible to achieve step function.

Cloud server preserves the initial data of user, and generates files on each of customers. To the aggregation of data analysis in a period of time, analyze the change of user movement situation.

Cell-phone customer terminal is different according to user and guardian role, has different authorities, and user in the gait situation of mobile phone terminal real time inspection oneself, can remind the incorrect attitude in user movement. Guardian can check location data, it is possible to receives the hazard condition warning applications of user.

Claims (5)

1. a gait analysis system for wearable Multi-information acquisition, its feature includes: flexible sensor module, 3-axis acceleration and gyro sensor, microprocessor, supply module, transport module, locating module, client and Cloud Server; It is made up of gait analysis device described 3-axis acceleration and gyro sensor, microprocessor, supply module and transport module and is arranged on the arch place of sole; Described client includes guardian's client and subscription client;

Described sensor assembly includes pliable pressure sensor (1) and soft stretch sensor (2);

Described pliable pressure sensor (1) is arranged on the heel portion of shoe pad, arch portion and forward foot in a step metacarpus, is used for gathering foot force information and passing to described microprocessor by adsorption-type pricking with needle (3);

Described soft stretch sensor (2) is arranged between the arch portion of shoe pad and forward foot in a step metacarpus, is used for gathering plantar flex degree and passing to described microprocessor also by adsorption-type pricking with needle (3);

Described 3-axis acceleration and gyro sensor are for gathering acceleration on three directions and angle of inclination and passing to described microprocessor;

Described locating module obtains the positional information of user and is sent to described microprocessor;

Received foot force information is calculated by described microprocessor, it is thus achieved that Center of Pressure dot information; And be analyzed according to described Center of Pressure dot information, plantar flex degree and acceleration, it is thus achieved that gait feature value;

Described gait feature value is input in neutral net and carries out gait judgement by described microprocessor, it is thus achieved that gait state result be sent to described client by described transport module, or be sent to described Cloud Server by described transport module and store; Make described client can obtain gait state result from described Cloud Server;

Described microprocessor is sent to described Cloud Server by described transport module after described positional information is processed and stores;

Described angle of inclination is analyzed by described microprocessor, it may be judged whether fall, if falling, then sends warning message to described Cloud Server by transport module; Described alarm information pushing is given described guardian's client by described Cloud Server again;

Described guardian's client obtains customer position information by described Cloud Server.

2. the gait analysis system of wearable Multi-information acquisition according to claim 1, is characterized in that: the composition of described pliable pressure sensor (1) including: flexible PCB (1a), copper electrode (1b), micro structure sensitive material (1c) and plastic sheeting (1d);

Flexible PCB (1a) is printed with described copper electrode (1b) on described, and covers described micro structure sensitive material (1c) on the surface of described copper electrode (1b); Described micro structure sensitive material (1c) the described plastic sheeting of upper covering (1d);

The composition of described soft stretch sensor (2) including: overlying plastic thin film (2a), conductive silver glue (2b), sensitive material (2c) and lower floor's plastic sheeting (2d);

Described lower floor plastic sheeting (2d) is printed with spaced two pieces of conductive silver glues (2b) as electrode; Described sensitive material (2c) is pasted onto in described lower floor plastic sheeting (2d) by described two pieces of conductive silver glues (2b); Overlying plastic thin film (2a) is covered on the surface of described sensitive material (2c).

3. the gait analysis system of wearable Multi-information acquisition according to claim 2, is characterized in that: described micro structure sensitive material (1c) is set to pyramid array format, and contacts with copper electrode (1b) with tower top.

4. the gait analysis system of wearable Multi-information acquisition according to claim 2, it is characterized in that: described micro structure sensitive material (1c) is after being mixed by rare to white carbon black and graphite mass ratio with 3:1, then be filled in silicone rubber molding with the total mass fraction of 4% and obtain;

Described sensitive material (2c) is after being mixed by rare to white carbon black and graphite mass ratio with 3:1, then is filled in silicone rubber molding with the total mass fraction of 6% and obtains.

5. a gait analysis method for wearable Multi-information acquisition, is characterized in that carrying out as follows:

Step 1, it is set to initial point O with a summit of the boundary rectangle of shoe pad, two adjacent sides of described initial point O is respectively set to X-axis and Y-axis, using the direction of vertical and described boundary rectangle as Z axis, constitute coordinate system O-XYZ;

Step 2, in described coordinate system XOY, obtain the position coordinates of n pliable pressure sensor, be designated as { (x₁,y₁),(x₂,y₂),��,(x_i,y_i),��,(x_n,y_n), (x_i,y_i) represent i-th pliable pressure sensor position coordinates; 1��i��n;

Step 3, utilize described n pliable pressure sensor to obtain n foot force value, be designated as { P₁,P₂,��,P_i,��,P_n; P_iRepresent the foot force value of i-th pliable pressure sensor;

Step 4, formula (1) and formula (2) is utilized to obtain Center of Pressure dot information (x_c,y_c):

$x_c=\underset{i=1}{\overset{n}{\Σ}}(x_i\×P_i)/\underset{i=1}{\overset{n}{\Σ}}{P}_i---\left(1\right)$

$y_c=\underset{i=1}{\overset{n}{\Σ}}(y_i\×P_i)/\underset{i=1}{\overset{n}{\Σ}}{P}_i---\left(2\right)$

Step 5, by soft stretch sensor (2) obtain t plantar flex degree C_t;

Step 6, obtain the acceleration on three directions of t by 3-axis acceleration and gyro sensorAnd utilize formula (3) to obtain the vector value S of t acceleration_t:

$S_t=\sqrt{{\left(a_x^t\right)}^2+{\left(a_y^t\right)}^2+{\left(a_z^t\right)}^2}---\left(3\right)$

When step 7, respectively collection normal gait and abnormal gait, Center of Pressure dot information (x_c,y_c), plantar flex degree C_t, acceleration on three directionsAnd the vector value S of acceleration_t, the respective seven kinds of eigenvalues of these seven kinds of data, for being trained neutral net, it is thus achieved that gait analysis model; Seven kinds of eigenvalues include maximum, minima, average, excursion, amplitude, variance and standard deviation;

Step 8, set described Center of Pressure dot information (x_c,y_c), plantar flex degree, acceleration on three directions and acceleration the respective seven kinds of eigenvalues of vector value walk accordingly threshold value;

Step 9, with T for the sampling period, F is sample frequency, formed sampling time window; In sampling time window, gather by Center of Pressure dot information (x_c,y_c), plantar flex degree, acceleration on three directions and acceleration the exercise data that forms of the respective seven kinds of eigenvalues of vector value;

Step 10, described exercise data and set walking threshold value are compared, it is judged that in sampling time window, whether user is at walking states; If at walking states, then perform step 11;

Step 11, described exercise data is inputted in described gait analysis model, thus obtaining the gait state result in sampling time window.