CN112245087B - An intelligent posture adjustment device - Google Patents

Abstract

The invention discloses an intelligent posture adjustment device, comprising a fixed body, a sensor, a state estimation module and a posture adjustment module. The sensors are respectively arranged on the abdomen contraction belt, shoulder contraction belt, waist contraction belt and back contraction belt. It is used to output the numerical data of the contact pressure between the abdominal contraction belt, shoulder contraction belt, waist contraction belt, and back contraction belt and the human body. The state estimation module is used to analyze the sensor data to obtain the characteristic information and posture data of the human body. The adjustment module is configured to adjust the body posture according to the output result of the state estimation module; the present invention adjusts the contraction belts of various parts in real time according to the user's state through the set state estimation module and body posture adjustment module, and then adjusts the various parts. The muscles exert force to achieve the correct posture and become more intelligent.

Description

An intelligent posture adjustment device

technical field

The invention relates to the technical field of intelligent wear, in particular to an intelligent body posture adjustment device.

Background technique

In the process of daily life and study, incorrect posture can cause many problems, and incorrect posture for a long time can lead to physical diseases. In the process of exercising, correct posture and correct force posture can save more effort and reduce muscle damage. In life, correct posture can improve personal temperament and image, reduce limb fatigue, spinal deformation and other health The problem is that in the growth and development stage, the correct posture can avoid problems such as hunchback and round shoulders. Many existing posture adjustment devices are generally mechanical restraints, which will cause physical discomfort when worn for a long time. The habit of time is used to correct the posture. This method is only suitable for the non-exercise state of the human body. When the human body is bending, leaning, shrugging, exercising, exercising, etc., the existing posture adjustment device cannot be adjusted according to different postures. , the limitation is relatively large. Therefore, the present invention proposes an intelligent body posture adjustment device that can solve the above problems.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide an intelligent body posture adjustment device to solve the above-mentioned problems in the background art.

In order to achieve the above purpose, the present invention provides the following technical solutions: an intelligent posture adjustment device, comprising:

The fixed main body is provided with a back contraction belt on the surface of the fixed main body. The fixed main body is divided into upper, middle and lower sections. The surface of the upper section of the fixed main body is connected with two shoulder contraction belts. The lower section surface is connected with a waist contraction belt, and the middle section surface of the fixed body is connected with an abdominal contraction belt;

Sensors, respectively arranged on the abdominal contraction belt, shoulder contraction belt, waist contraction belt, and back contraction belt, are used to output the numerical data of the contact pressure between the abdominal contraction belt, the shoulder contraction belt, the waist contraction belt, and the back contraction belt and the human body ;

The state estimation module is set on the fixed body, and is used to analyze the sensor data to obtain the characteristic information and posture data of the human body;

The body posture adjustment module is arranged on the fixed main body, and is used to adjust the body posture according to the output result of the state estimation module.

As a preferred technical solution of the present invention, the fixing body is made of hard leather, and the upper, middle and lower sections of the fixing body are connected with each other by a flexible material.

As a preferred technical solution of the present invention, control motors are provided on the upper, middle and lower sections of the fixed body, and the control motors are respectively connected with the abdominal contraction belt, shoulder contraction belt, waist contraction belt, and back contraction belt. Belt drive connection, the fixed body is a multi-layer structure, the control motor is placed in the interlayer of the fixed body, the sensor is located in the abdomen contraction belt, shoulder contraction belt, waist contraction belt, back contraction belt and the control motor are connected At one end, the other ends of the abdominal contraction belt, the shoulder contraction belt and the waist contraction belt are bonded to each other.

As a preferred technical solution of the present invention, hooks are provided at the contacting parts of the abdominal contraction belt, the shoulder contraction belt, and the waist contraction belt.

As a preferred technical solution of the present invention, the abdominal contraction belt, shoulder contraction belt, waist contraction belt and back contraction belt are elastic straps.

As a preferred technical solution of the present invention, the state estimation module includes a neural network unit and a deviation data output module. The neural network unit is used for analyzing the value output by the sensor to obtain the type of motion, and the deviation data output module is used for according to Compare the type data of the exercise in the database and output the deviation of the pressure of each muscle exerting part from the standard amount.

As a preferred technical solution of the present invention, the posture adjustment module adjusts the abdomen contraction belt, shoulder contraction belt, waist contraction belt, and back contraction belt by controlling the motor.

As a preferred technical solution of the present invention, the body posture adjustment module includes a wireless communication module, and the wireless communication module uploads data to the host computer for displaying body posture characteristics and correct posture in a daily time period.

Compared with the prior art, the beneficial effects of the present invention are:

According to the user's state, the present invention adjusts the contraction belt of each part in real time through the set state estimation module and posture adjustment module, and then adjusts the muscle force of each part to achieve the correct posture, which can be adjusted according to the different postures of the user. adjustment, more intelligent.

Description of drawings

Fig. 1 is the overall structure schematic diagram of the present invention;

Fig. 2 is the overall side view structure schematic diagram of the present invention;

3 is a system block diagram of the present invention;

Fig. 4 is the general flow chart of the present invention;

Fig. 5 is the flow chart of the state estimation of the present invention;

Fig. 6 is the flow chart of posture adjustment of the present invention;

In the figure: 101, fixed body; 102, abdominal contraction belt; 103, shoulder contraction belt; 104, waist contraction belt; 105, hook; 107, control motor; 108, sensor; 109, back contraction belt; 110, flexible material .

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

In the description of the present invention, it should be noted that the orientations or positional relationships indicated by the terms "vertical", "upper", "lower", "horizontal", etc. are based on the orientations or positional relationships shown in the drawings, only In order to facilitate the description of the present invention and simplify the description, it is not indicated or implied that the indicated device or element must have a particular orientation, be constructed and operated in a particular orientation, and therefore should not be construed as limiting the present invention.

In the description of the present invention, it should also be noted that, unless otherwise expressly specified and limited, the terms "arranged", "installed", "connected" and "connected" should be understood in a broad sense, for example, it may be a fixed connection, It can also be a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or the internal communication between the two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

1-6, the present invention provides a technical solution: an intelligent posture adjustment device, including a fixed body 101, a sensor 108, a state estimation module, and a posture adjustment module, and the surface of the fixed body 101 is provided with a back shrinking belt 109. The fixed main body 101 is divided into upper, middle and lower sections. The upper surface of the fixed main body 101 is connected with two shoulder contraction belts 103, the lower surface of the fixed main body 101 is connected with waist contraction belts 104, and the middle surface of the fixed main body 101 is connected with a waist contraction belt 104. There is an abdominal contraction belt 102 connected to it, the fixed body 101 and the two shoulder contraction belts 103 control the balance of the two shoulders, the fixed main body 101 and the abdominal contraction belt 102 control the upper abdomen, and the upper abdomen does not have the action of human body movement, so this In order to balance the base point and control the adjustment of the shoulders and waist, the fixed body 101 and the waist contraction belt 104 are used to tighten the abdomen, and the abdominal contraction belt 102 is used to adjust the back of the human body. Fit the lumbar spine and keep the central axis of the fixed body 101 in the same position as the central axis of the waist of the human body. Since the body is self-wearing, the central axis cannot be guaranteed to be completely consistent, but the probability of incorrect posture of the chest of the human body is small. It is achieved that the fixed body 101 and the upper body are correctly worn, and are nearly consistent with the central axis.

The sensors 108 are respectively arranged on the abdominal contraction belt 102, the shoulder contraction belt 103, the waist contraction belt 104, and the back contraction belt 109. Two sensors 108 are symmetrically arranged in front of and behind the body to improve the sensitivity of the body posture, and are used to output the abdominal contraction belt 102. Numerical data of the contact pressure of the shoulder contraction belt 103, the waist contraction belt 104, and the back contraction belt 109 with the human body.

其值为正说明肌肉收缩正在使用，由于传感器108输出的是数字量，所以采用进行

N为数据的个数，对两个对称的传感器108的数值x_H和x_k进行差比和的方式得到响应数据

各传感器108组成数据矩阵

d₁，d₂......d₉对应从左到右，从上到下的对称传感器108差比和数据，神经网络单元内含有信息处理模块对一系列的数据矩阵A_t计算出运动类型T_x，偏差数据输出模块根据矩阵

和运动类型T_x对比数据库中运动的类型数据输出传感器108部位与标准量的偏差C_x，构建偏差矩阵

传入下个模块进行使用，该系统首先需要通过大量人体状态和身姿状态数据对神经网络需要进行训练，得出正确的概率模型，由于人体运动是连续的，传感器的受压力也是连续变化的，运动状态与标准数值不存在绝对相同，所以以概率给出。每种运动类型的概率为

概率总和

为条件概率，当概率超过设定的阈值则判定用户当前是此运动，如果全部的类型概率都不满足则返回类型不存在的标志，该神经网络的实例模型CNN+LSTM，在检测进行正确穿戴后，每秒各部位传感器108数据拼接成矩阵进行分析，已经训练成功的神经网络直接输出用户身姿信息，身姿信息再与数据库存在的类型进行比对，如果为数据库存在的身姿，偏差数据输出模块则输出与标准身姿的相差数据量，否则返回不存在类型，该状态推测模块的时间分辨率为一秒。The state estimation module is arranged on the fixed body 101, and is used to analyze the data of the sensor 108 to obtain the characteristic information and posture data of the human body. The state estimation module includes a neural network unit and a deviation data output module. It is going on every moment, but there is a relatively static situation, so the data input by the state estimation module is multiple data matrices of time difference, any sensor 108 is x, and x _t is the pressure received by the pressure sensor 108, which is very short. Averaging over time

A positive value indicates that the muscle contraction is in use. Since the output of the sensor 108 is a digital quantity, the

N is the number of data, and the response data is obtained by performing the difference ratio and sum of the values x _H and x _k of the two symmetrical sensors 108

Each sensor 108 forms a data matrix

d ₁ , d ₂ ...... d ₉ correspond to the difference ratio and data of the symmetrical sensor 108 from left to right, from top to bottom, the neural network unit contains an information processing module to calculate a series of data matrix A _t Motion type T _x , deviation data output module according to matrix

Compare the motion type data in the database with the motion type T _x and output the deviation C _x of the sensor 108 position from the standard quantity, and construct a deviation matrix

Pass it into the next module for use. The system first needs to train the neural network through a large amount of human body state and posture state data to obtain the correct probability model. Since the human body movement is continuous, the pressure on the sensor also changes continuously. , the motion state does not exist absolutely the same as the standard value, so it is given by probability. The probability of each exercise type is

sum of probabilities

It is a conditional probability. When the probability exceeds the set threshold, it is determined that the user is currently in this movement. If all the types of probabilities are not satisfied, the flag that the type does not exist will be returned. The instance model of the neural network CNN+LSTM is correctly worn during detection. After that, the data of each part of the sensor 108 is spliced into a matrix for analysis every second, the neural network that has been successfully trained directly outputs the user's posture information, and the posture information is compared with the type that exists in the database. If it is the posture that exists in the database, the deviation The data output module outputs the difference data amount from the standard posture, otherwise it returns the non-existent type, and the time resolution of the state inference module is one second.

The fixed main body 101 is made of hard leather material. The upper, middle and lower sections of the fixed main body 101 are provided with flexible materials 110 to connect with each other. The hard leather material has certain toughness and softness. The hard leather material makes the upper, middle and lower sections of the fixing body 101 have a certain hardness and is not easy to bend, and the flexible material 110 between each section can be folded, reducing the volume and making it easy to carry.

The upper, middle and lower sections of the fixed main body 101 are provided with control motors 107, and the control motors 107 are respectively connected to the abdomen contraction belt 102, the shoulder contraction belt 103, the waist contraction belt 104, and the back contraction belt 109. It is a multi-layer structure, the control motor 107 is placed in the interlayer of the fixed main body 101, the sensor 108 is located at the end of the abdomen contraction belt 102, the shoulder contraction belt 103, the waist contraction belt 104, the back contraction belt 109 and the control motor 107, the abdomen is contracted. The other ends of the belt 102 , the shoulder shrinking belt 103 , and the waist shrinking belt 104 are bonded to each other, and circuit wires are wrapped in the interlayer of the fixing body 101 .

The posture adjustment module is arranged on the fixed main body 101, and is used to adjust the posture according to the output result of the state estimation module. The posture adjustment module controls the motor 107 to adjust the abdominal contraction belt 102, the shoulder contraction belt 103, and the waist contraction belt 104. , The back contraction belt 109 is adjusted, and the posture adjustment module has a feedback adjustment mechanism. When the user's state and data are obtained according to the current sensor 108 data, if it is exercise, the contraction belt of each part will be expanded and contracted to improve the convenience of exercise. , when the obtained state is non-movement, stretch the contraction belt of each part, control the pulling force and speed of the motor 107 according to the tightening degree fed back by the sensor 108, and restrain the posture to different degrees. When the user is in a state of motion When the user is in an unconstrained state, adjust which part of the muscle to exert force to improve its sensitivity and comfort. When the user is standing or sitting, slowly stretch the contraction belt, and infer the data output from the module according to the state. Stretch each part of the contraction belt to achieve a standard posture.

The control motor 107 has a suitable torque and rotates slowly, and the control motor 107 has the function of reversing. Since the human body is in a restrained state, it will not make particularly violent movements. In this embodiment, when the human body encounters a special situation, it needs a large The range changes the posture, the control motor 107 can be reversed to a certain extent through the gears, and the generated reverse voltage needs to be isolated by the isolation element. After the reversal, the positive feedback adjustment is performed, and the human body is not bound. When the human body is stable, the posture is performed again. Adjustment.

的数据则表示各连接电机需要进行拉伸的数值量，当C_x为正值则表示该传感器的受到压力比标准值大，需要拉伸电机，使身体挺立，减少此部位用力。其中调节采用动态差值调节ΔU_t＝K_p(C_t-C_t-1)+K_i(C_t)+K_d(C_t-2*C_t-1+C_t-2)，采用增量式PID有一定的惯性缓冲量，减少电机的立刻控制对身体的伤害，身姿调整模块最后结果是根据标准偏差数据量进行动态调整，达到一个动态平衡，增加装置的适应性和便利性，该过程为首先判断用户的状态，如果是运动状态，针对不同的运动，跑步、跳跃、跳绳等做不同的身姿调整，先放开各部位的收缩带，使人体能更好的活动关节，再根据与标准身姿的相差数据量，控制电机107调整各部位收缩带，不断减小这个相差值，调控的程度有一定的范围限度，即限制调控输出最大值，会一定程度上延长调整的调节时间，调整具体稳定性和舒适性，多次反馈调节后，达到微小的相差量，达到动态平衡，如果返回不存在类型，再进一步根据用户的运动范围程度进行正反馈，达到不妨碍用户运动。The input of the posture adjustment module is the output data of the state inference module, and the matrix

The data of , indicates the value of each connected motor that needs to be stretched. When C _x is a positive value, it means that the pressure of the sensor is greater than the standard value, and the motor needs to be stretched to make the body stand upright and reduce the force on this part. The adjustment adopts dynamic difference adjustment ΔU _t =K _p (C _t -C _t-1 )+K _i (C _t )+K _d (C _t -2*C _t-1 +C _t-2 ), using increasing The quantitative PID has a certain amount of inertia buffer, which reduces the damage to the body caused by the immediate control of the motor. The final result of the posture adjustment module is to dynamically adjust according to the standard deviation data volume to achieve a dynamic balance and increase the adaptability and convenience of the device. The process is to first judge the state of the user. If it is a sports state, make different posture adjustments for different sports, such as running, jumping, skipping rope, etc., first release the contraction belt of each part, so that the human body can better move the joints. Then, according to the difference data amount from the standard posture, the motor 107 is controlled to adjust the contraction belt of each part, and the difference value is continuously reduced. Adjust the time, adjust the specific stability and comfort, after several feedback adjustments, a small difference is achieved, and dynamic balance is achieved. If there is no type of return, further positive feedback will be carried out according to the user's range of motion, so as not to hinder the user's movement. .

The posture adjustment module includes a wireless communication module, and the wireless communication module uploads data to the host computer, which is used to display the posture characteristics and correct posture of the daily time period. In this embodiment, the wireless communication module is realized by Bluetooth, and the host computer sends the specified protocol data to Posture adjustment module, the adjustment module gets the command command, if it is to modify the parameter command of the adjustment module, it will return the modification result, if it is to read the data file, return the data stored by the adjustment module to the upper computer, and the user can see the data within the time period. The body posture information and body posture accuracy rate realize the interactive function of the system.

Hooks 105 are provided at the parts where the abdomen contraction belt 102, the shoulder contraction belt 103, and the waist contraction belt 104 are in contact with each other.

The abdominal contraction belt 102 , the shoulder contraction belt 103 , the waist contraction belt 104 and the back contraction belt 109 are elastic straps to prevent the contraction belts in various parts from causing strangulation to the body.

Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, and substitutions can be made in these embodiments without departing from the principle and spirit of the invention and modifications, the scope of the invention is defined by the appended claims and their equivalents.

Claims (8)

1. an intelligent posture adjustment device, is characterized in that, comprises: A fixed main body (101), the surface of the fixed main body (101) is provided with a back contraction belt (109), the fixed main body (101) is divided into upper, middle and lower sections, the upper section of the fixed main body (101) Two shoulder contraction belts (103) are connected to the surface, a waist contraction belt (104) is connected to the lower surface of the fixed main body (101), and an abdominal contraction belt (102) is connected to the middle surface of the fixed main body (101). ); Sensors (108) are respectively arranged on the abdominal contraction belt (102), the shoulder contraction belt (103), the waist contraction belt (104), and the back contraction belt (109), and are used for outputting the abdominal contraction belt (102), the shoulder belt Numerical data of the contact pressure of the contraction belt (103), waist contraction belt (104), and back contraction belt (109) with the human body; The state estimation module is arranged on the fixed main body (101), and is used for analyzing the sensor data to obtain the characteristic information and posture data of the human body; the state estimation module includes a neural network unit and a deviation data output module. It is going on all the time, but there is a relatively static situation, so the data input by the state estimation module is multiple data matrices of time difference, any sensor is x, and x _t is the pressure received by the pressure sensor, and averaged deal with

Its value is positive, indicating that the muscle contraction is in use. Since the sensor outputs a digital quantity, using

N is the number of data, and the response data is obtained by comparing and summing the values x _H and x _k of two symmetrical sensors.

Each sensor forms a data matrix

d ₁ , d ₂ ...... d ₉ correspond to the symmetric sensor difference ratio and data from left to right, from top to bottom, the neural network unit contains an information processing module to calculate the motion for a series of data matrices A _t Type T _x , the deviation data output module according to the matrix

Compare the motion type data in the database with the motion type T _x , output the deviation C _x of the sensor position from the standard quantity, and construct a deviation matrix

To pass it into the next module for use, first, the neural network needs to be trained through a large amount of human body state and posture state data to obtain the correct probability model. The probability of each movement type is

sum of probabilities

It is a conditional probability. When the probability exceeds the set threshold, it is determined that the user is currently in this movement. If all the types of probabilities are not satisfied, the flag that the type does not exist will be returned. The instance model of the neural network CNN+LSTM is correctly worn during detection. After each second, the sensor data of each part is spliced into a matrix for analysis. The successfully trained neural network directly outputs the user's posture information, and the posture information is compared with the type existing in the database. If it is the posture that exists in the database, the deviation data The output module outputs the difference data amount from the standard posture, otherwise it returns the non-existent type, and the time resolution of the state inference module is one second; The body posture adjustment module is arranged on the fixed main body (101), and is used for adjusting the body posture according to the result output by the state estimation module. 2. An intelligent posture adjustment device according to claim 1, characterized in that, the fixed body (101) is made of hard leather, and the upper, middle and lower sections of the fixed body (101) are provided with The flexible materials (110) are interconnected. 3. An intelligent posture adjustment device according to claim 1, characterized in that a control motor (107) is provided on the upper, middle and lower sections of the fixed body (101), and the control motor (107) are respectively connected to the abdomen contraction belt (102), the shoulder contraction belt (103), the waist contraction belt (104), and the back contraction belt (109). The control motor (107) is placed in the interlayer of the fixed body (101), and the sensors (108) are located in the abdominal contraction belt (102), the shoulder contraction belt (103), the waist contraction belt (104), the back contraction belt (109) ) is connected with the control motor (107), and the other ends of the abdominal contraction belt (102), the shoulder contraction belt (103), and the waist contraction belt (104) are bonded to each other. 4. An intelligent posture adjustment device according to claim 1, characterized in that, the contacting parts of the abdomen contraction belt (102), the shoulder contraction belt (103), and the waist contraction belt (104) are provided with Hook (105). 5. An intelligent posture adjustment device according to claim 1, wherein the abdomen contraction belt (102), the shoulder contraction belt (103), the waist contraction belt (104) and the back contraction belt (109) ) are elastic straps. 6. An intelligent posture adjustment device according to claim 1, characterized in that, the state estimation module comprises a neural network unit and a deviation data output module, and the neural network unit is used to perform a calculation on the value output by the sensor (108). The type of exercise is obtained by analysis, and the deviation data output module is used to output the deviation between the pressure of each muscle exerting part and the standard amount according to the type of exercise data in the comparison database. 7. An intelligent posture adjustment device according to claim 1, characterized in that, the posture adjustment module controls the motor (107) to adjust the abdominal contraction belt (102), the shoulder contraction belt (103), the waist The shrink belt (104) and the back shrink belt (109) are adjusted. 8 . An intelligent posture adjustment device according to claim 1 , wherein the posture adjustment module comprises a wireless communication module, and the wireless communication module uploads data to the host computer for displaying the posture characteristics and Correct posture.

Images