CN108926813B - Training method based on human body balance data - Google Patents

Abstract

The invention discloses a training method based on human body balance data, and belongs to the field of rehabilitation training methods. The training method comprises the steps of firstly carrying out real-time data acquisition on an individual under eight preset interactive posture balance tasks, then analyzing the acquired data by using a nonlinear dynamics method to obtain the quantitative values of the overall balance function of the individual and the limb coordination capacity of the individual, and finally comparing the quantitative values with indexes in a database to obtain the maximum angles and the corresponding maximum time for keeping balance of the eight postures of the human body, thereby realizing the balance training of the eight postures of the human body. The invention can complete the evaluation of the balance function and the corresponding training only by the intelligent mobile phone carried by the human body, is not limited by time and space, and can test and train the balance function of the human body for a long time.

Description

Training method based on human balance data

technical field

The invention relates to a training method based on human body balance data, and belongs to the field of rehabilitation training methods.

Background technique

Balance is an important function of the human body, and the ability to balance in daily life is especially important for maintaining various postures, performing various activities, and responding appropriately to external disturbances.

At present, there are many methods for evaluating and training balance function. Among them, the traditional subjective observation method is simple and easy to operate, convenient, intuitive and fast, but it is too rough and subjective and lacks objective quantitative standards. It can only be used for the preliminary screening of patients suspected of having balance dysfunction. check. The scale method is easy to quantify and compare, but the operation is cumbersome and time-consuming, and it is affected by human factors, and the error is high. The pressure plate test is simple and quick to operate, but highly specialized and expensive, and is only suitable for research and application. These assessments and tests make it difficult for the general public to pay enough attention to the balance function, let alone improve it scientifically and effectively. Therefore, a method that is convenient for the public to objectively evaluate and scientifically improve the self-balancing function in a conventional environment at any time needs to be proposed urgently.

SUMMARY OF THE INVENTION

In order to solve the current lack of training methods for the balance function of the human body, the present invention provides a training method based on the balance data of the human body, which can evaluate the balance function for the public anytime and anywhere, and provide a training implementation plan.

The technical scheme of the present invention is as follows:

A training method based on human body balance data, comprising the following steps:

1) When the lower body does not move, the upper body tilts forward, backward, left, right, left forward, left backward, right forward, and right backward. The acceleration sensor and angular velocity sensor are used to collect 8 groups of postures of the human body. real-time data of acceleration and angular velocity signals;

Specifically, the sensor built in the mobile phone is used to collect the real-time data of the acceleration and angular velocity signals of the human body under the interactive attitude balance task. There are 8 kinds of interactive posture balance tasks, including: when the individual's lower body is not moving, the upper body leans forward, backward, left, right, left forward, left backward, right forward, and right backward; interactive The method is to prompt the individual to complete the next preset posture balance task through mobile phone voice and flashing. When the current posture balance task is completed, the mobile phone will send out a prompt tone; the collection time is 1 minute to 15 minutes, and the sampling frequency is 100Hz to 300Hz.

2) The real-time data of 8 groups of human body poses are reconstructed in phase space through delay time τ and embedding dimension E, and 8 groups of E-dimensional time series Z _i (i=1, 2,..., E) are obtained, and further use The local prediction method predicts each other in phase space of these 8 groups of time series, and the similarity degree of prediction is used as the quantitative value of limb coordination ability;

其中

为原始加速度数据中相邻两个数据的最大差值；将归一化的样本熵加权就作为人体整体平衡功能定量值；3) Perform multi-scale entropy analysis on 8 groups of E-dimensional time series Z _i (i=1, 2, ..., E) obtained by phase space reconstruction, and obtain 8 sample entropy

in

is the maximum difference between two adjacent data in the original acceleration data; the normalized sample entropy is weighted as the quantitative value of the overall balance function of the human body;

4) Compare the quantitative value of the above-mentioned limb coordination ability and the quantitative value of the overall balance function of the human body with the indexes of the database respectively, and obtain the maximum angle at which the 8 postures of the human body maintain balance and the corresponding maximum time;

5) The angle and corresponding time for maintaining balance in each posture of the human body are greater than the maximum angle and the corresponding maximum time obtained above, and 8 kinds of posture balance training of the human body are realized.

Among them, the database contains ten levels of data sets of the overall balance ability and limb coordination ability of the population, and each level is divided into 1-4 layers from small to large; the database records the individual in each level of each level. The maximum angle θ _ijk (i=1, 2,...8; j=1, 2,...10; k=1, 2, 3, 4) for maintaining balance under the attitude task and the corresponding maximum time T _ijk (i=1,2,...8; j=1,2,...10; k=1,2,3,4), where i represents 8 pose tasks, j represents ten levels, k Represents four levels.

The training volume is determined by the maximum angle and the corresponding maximum time for the human body to maintain balance under 8 posture tasks, so as to improve the balance function of the human body. Among them, the maximum angle of the training amount to maintain balance is between θ _ijk and θ _ij(k+1) , and the corresponding maximum time is between T _ijk and T _ij(k+1) .

The present invention has the following advantages:

The present invention is not limited by time and space, and only uses the smart phone carried by the individual without purchasing any other equipment, so that the evaluation and training of the balance function of the human body can be performed anytime and anywhere, and long-term detection can be provided; Personalized training programs can scientifically and effectively improve one's own balance function.

Description of drawings

Fig. 1 is the schematic flow chart of the method of the present invention;

Fig. 2 is the schematic diagram of 8 kinds of attitude tasks of individuals in the method of the present invention;

FIG. 3 is a flow chart of balanced training in the method of the present invention.

Detailed ways

The present invention will be further described below through specific embodiments for better understanding of the present invention, but the present invention is not limited thereto.

1 is a schematic flow chart of the method of the present invention, which is mainly divided into data collection, calculation of the quantitative value of the overall balance function of the human body and the quantitative value of the limb coordination ability, database comparison and training plan formulation. The whole system can use Bayesian estimation to form a closed loop to continuously improve the individual balance ability.

First, establish the basic file of the individual according to the input name, gender, age, height, weight and medical history.

When evaluating the individual balance function, the acceleration and angular velocity sensors built in the smartphone are used to collect the acceleration and angular velocity signals of the human body in real time under eight preset interactive posture balance tasks. The acquisition time is 1 minute to 15 minutes. The sampling frequency is adjustable from 100Hz to 300Hz, and is divided into the following steps:

(1) After clicking to start the test on the mobile phone application interface, the individual's body is naturally upright, his feet are together, his arms are crossed at the chest, and the mobile phone is supported with both hands and the screen is placed on the chest;

(2) As shown in Figure 2, the individual keeps the body below the waist still, and the upper body randomly moves forward 1, backward 2, left 3, right 4, left forward 5, left backward 6, and forward according to the mobile phone prompt tone. The front right 7 and the rear right 8 are tilted in eight directions. After each tilt, when you hear a beep sound from the phone, it will return to the upright state and prepare for the next action;

(3) After hearing the notification from the mobile phone that the test has been completed, remove the mobile phone from the chest and click to save the data.

其中

为原始加速度数据中相邻两个数据的最大差值；将归一化的样本熵加权就作为人体的整体平衡功能定量值，其中权重为各个姿态平衡任务下采集数据的8个最大李雅普诺夫指数

The real-time data collected under 8 preset posture balance tasks when the human body is standing is reconstructed in phase space through the delay time τ and the embedding dimension E, and 8 groups of E-dimension time series Z _i (i=1, 2, . .., E), further use the local prediction method to predict each other in the phase space of the 8 groups of time series, and the similarity of the predictions is used as the quantitative value of the limb coordination ability. Perform multi-scale entropy analysis on 8 groups of E-dimensional time series Z _i (i=1, 2, ..., E) obtained by phase space reconstruction, and obtain 8 sample entropy

in

is the maximum difference between two adjacent data in the original acceleration data; the normalized sample entropy weight is used as the quantitative value of the overall balance function of the human body, where the weight is the 8 largest Lyapunov values of the data collected under each attitude balance task index

Compare the calculated quantitative value of the individual's overall balance function and the quantitative value of limb coordination with the indicators in the database, and obtain the maximum angle and the corresponding maximum time for the human body to maintain balance under the eight preset posture balance tasks. ; Among them, the human body balance function and limb coordination ability database contains ten levels of overall balance ability and limb coordination ability data sets, and each level is divided into 1-4 layers from small to large; the human body balance function and limb coordination ability database The maximum angle θ _ijk (i=1,2,...8; j=1,2,...10; k=1,2, 3,4) and the corresponding maximum time T _ijk (i=1,2,...8; j=1,2,...10; k=1,2,3,4), where i,j , k represent eight posture tasks, ten levels, and four levels, respectively.

As shown in Figure 3, the training volume is selected according to the overall balance function. The training volume refers to the maximum angle θ _ijk of the previous training and the longest time T _ijk corresponding to maintaining balance, and the training volume increases by one level each time, that is, two levels 1/4 of the difference. If the previous training amount is (θ _ijk , T _ijk ), the current training amount is [θ _ijk +0.25(θ _i(j+1)k -θ _ijk ), T _ijk +0.25(T _{i(j+ 1)k} -T _ijk )], the next training amount is [θ _ijk +0.50(θ _i(j+1)k -θ _ijk ), T _ijk +0.50(T _i(j+1)k -T _ijk )], until the training volume reaches (θ _i(j+1)k , T _i(j+1)k ), then re-evaluate the individual’s overall balance function and limb coordination ability, and update the training program.

则每种交互式姿态平衡任务在下次训练中出现的比例为(θ_sum-θ_i)/θ_sum。The number of eight training tasks for posture balance is proportionally selected according to the degree of the individual's insufficient limb coordination function, and the sum of the typical inclination angles θ _i under the preset eight interactive posture balance tasks is set to be θ _sum ,

Then the proportion of each interactive pose balance task in the next training is (θ _sum -θ _i )/θ _sum .

Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art, without departing from the scope of the technical solution of the present invention, can make many possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above, or modify it into an equivalent implementation of equivalent changes. example. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solutions of the present invention still fall within the protection scope of the technical solutions of the present invention.

Claims (9)

1. A training method based on human body balance data, comprising the following steps: 1) When the lower body does not move, the upper body tilts forward, backward, left, right, left forward, left backward, right forward, and right backward. The acceleration sensor and angular velocity sensor are used to collect 8 groups of postures of the human body. real-time data of acceleration and angular velocity signals; 2) The real-time data of 8 groups of human body poses are reconstructed in phase space through delay time τ and embedding dimension E, and 8 groups of E-dimensional time series Z _i (i=1, 2,..., E) are obtained, and further use The local prediction method predicts each other in phase space of these 8 groups of time series, and the similarity degree of prediction is used as the quantitative value of limb coordination ability; 3) Perform multi-scale entropy analysis on 8 groups of E-dimensional time series Z _i (i=1, 2, ..., E) obtained by phase space reconstruction, and obtain 8 sample entropy

in

is the maximum difference between two adjacent data in the original acceleration data; the normalized sample entropy is weighted as the quantitative value of the overall balance function of the human body; 4) Compare the quantitative value of the above-mentioned limb coordination ability and the quantitative value of the overall balance function of the human body with the indexes of the database respectively, and obtain the maximum angle at which the 8 postures of the human body maintain balance and the corresponding maximum time; 5) The angle and corresponding time for maintaining balance in each posture of the human body are greater than the maximum angle and the corresponding maximum time obtained above, and 8 kinds of posture balance training of the human body are realized. 2 . The training method according to claim 1 , wherein step 1) comprises: collecting real-time data by using a mobile phone with an acceleration sensor and an angular velocity sensor set therein. 3 . 3. The training method according to claim 2, wherein the collecting step comprises: 1) After clicking to start the test on the mobile phone application interface, the individual's body is naturally upright, his feet are together, his arms are crossed at the chest, and the mobile phone is supported with both hands and the screen is placed on the chest; 2) The individual keeps the body below the waist still, and the upper body randomly moves forward 1, back 2, left 3, right 4, left front 5, left back 6, right front 7 and right back 8 according to the mobile phone beep A total of eight directions are tilted. After each tilt, when you hear a beep sound from the phone, it will return to the upright state and prepare for the next action; 3) After hearing the prompt from the mobile phone that the test has been completed, remove the mobile phone from the chest and click to save the data. 4 . The training method according to claim 3 , wherein the time for each inclination is 1 minute to 15 minutes, and the sampling frequency is 100 Hz to 300 Hz. 5 . 5. training method as claimed in claim 1 is characterized in that, in step 3), weight is 8 maximum Lyapunov exponents that collect data under each attitude balance task

6. The training method according to claim 1, wherein the index of the database in step 4) includes ten levels, and each level is divided into 1-4 layers from small to large. 7. The training method according to claim 6, wherein the maximum angle θ _ijk (i=1, 2, . . . 8; j=1, 2, ... 10; k = 1, 2, 3, 4) and the corresponding maximum time T _ijk (i = 1, 2, ... 8; j = 1, 2, ... 10; k = 1, 2, 3, 4), where i represents eight pose tasks, j represents ten levels, and k represents four levels. 8. training method as claimed in claim 7 is characterized in that, the angle that the human body keeps balance under 8 kinds of posture tasks satisfies between θ _ijk to θ _{ij (k+1)} , and the time that keeps balance satisfies between T _ijk to θ ijk between T _ij(k+1) . 9. training method as claimed in claim 7 is characterized in that, the sum of the angle θ _i that the human body maintains balance under posture balance task is θ _sum ,

The proportion of each pose balance task that occurs during training is (θ _sum -θ _i )/θ _sum .

Images