JP2016041155A - Posture and walking state estimation device - Google Patents

Abstract

A posture and walking state estimation device capable of estimating a walking state and / or posture more easily and accurately. A detection means 10 attached to a human body and an analysis means 20 for analyzing a walking state and / or posture based on a signal from the detection means 10 are provided. The vertical acceleration detection unit 12 that detects the vertical acceleration of the body of the human body, the analysis means 20 has a posture stability determination function 22 that determines the stability of the posture, and the posture stability determination function 22 The floor reaction force is estimated based on the signal from the vertical acceleration detection unit 12, and the walking state and / or posture is determined based on the time variation of the estimated floor reaction force. [Selection] Figure 1

Description

  The present invention relates to a posture and walking state estimation device. More specifically, the present invention relates to a posture and a walking state estimation device capable of estimating a walking state while being worn by a person.

  2. Description of the Related Art Conventionally, techniques for processing a signal from an angular velocity sensor or an acceleration sensor attached to a human body to estimate a human walking state have been developed (see Patent Documents 1 to 4).

Patent Document 1 discloses a technique for calculating a walking speed by calculating a stride and a cycle from an acceleration curve in the front-rear direction. Also disclosed is a technique for grasping the degree of resolution of lower limb dysfunction by grasping the left and right stances from the power spectrum of the acceleration waveform in the left and right direction.
Patent Document 2 discloses a technique for comparing a gait and a normal person's walking with a difference in frequency of each acceleration around a horizontal axis.
Furthermore, in Patent Document 3, an acceleration sensor is attached to the waist, and the difference between the peak and valley peak of the vertical direction acceleration component and the difference between the peak and valley peak of the walking direction acceleration component are calculated and prepared in advance. A technique for estimating a walking speed and a stride as a walking ability by a relational expression between a walking speed and a walking speed is disclosed.
Patent Document 4 discloses a technique for evaluating dorsiflexion force, lower limb muscle strength, and the like simply by attaching an accelerometer to the waist and measuring the acceleration of the waist. There is also a statement that ability can be estimated.

JP 2004-358229 A Japanese Patent Laid-Open No. 2003-6608 JP 2005-114537 A JP 2007-125368 A

  Even in the techniques of Patent Documents 1 to 4, a certain walking state or the like can be detected by attaching a sensor or the like to a human body. However, even in the techniques of Patent Documents 1 to 4, estimation of the walking state is insufficient, and an apparatus that can estimate the walking state more easily and accurately is demanded.

  An object of this invention is to provide the attitude | position and walking state estimation apparatus which can estimate a walking state and / or attitude | position more simply and accurately in view of the said situation.

The posture and walking state estimation device of the first invention includes a detecting means attached to a human body and an analyzing means for analyzing the walking state and / or posture based on a signal from the detecting means, The detection means includes a vertical acceleration detection unit that detects vertical acceleration of a human body, and the analysis means has a posture stability determination function that determines posture stability. The stability determination function has a function of estimating a vertical floor reaction force based on a signal from the vertical acceleration detection unit and determining a walking state and / or posture based on a temporal variation of the estimated vertical floor reaction force. It is characterized by that.
In the posture and walking state estimation device according to a second aspect of the present invention, in the first aspect, the posture stability determination function is based on a relative position between the detection means and the position of the center of gravity of the human body, and a signal from the detection means. On the basis of this, the change in the position of the center of gravity is grasped, and the stability of the walking state and / or posture of the person is determined based on the temporal variation of the position of the center of gravity.
In the posture and walking state estimation device according to a third aspect of the present invention, in the first or second aspect, the analysis means estimates an inertial force applied to a center of gravity position of a human body, and calculates a center of gravity position in a state where the inertial force is applied. The present invention is characterized in that the stability of a person's walking state and / or posture is determined based on time variation.
The posture and walking state estimation device according to a fourth aspect of the present invention is the first, second or third aspect, wherein the posture stability determination function is based on a relative positional relationship between the gravity center position and the pressure center. It is characterized by determining the stability of the state and / or posture.
The posture and walking state estimation device according to a fifth aspect of the present invention is the first, second, third, or fourth aspect, wherein the detection means includes a longitudinal acceleration detection unit that detects the longitudinal acceleration of the human body. The analyzing means has a function of determining a whirl based on a signal from the longitudinal acceleration detector and a temporal variation of the floor reaction force.

According to the first aspect of the invention, since the time variation of the vertical floor reaction force is estimated, it is possible to determine the stability of the person's walking state and / or posture simply and with a certain degree of accuracy.
According to the second invention, since it is possible to grasp a change or abnormality in the walking state or posture of a person, it is possible to determine a physical condition or a malfunction of the subject (for example, an injury) from an abnormality in the walking or posture of the subject. There is a possibility.
According to the third aspect of the invention, the stability of the person's walking state and / or posture is determined from the inertial force applied to the position of the center of gravity of the person's body and the temporal variation of the position of the center of gravity. The stability of the posture can be determined.
According to the fourth aspect of the present invention, the stability of the person's walking state and / or posture is determined from the relative position of the center of gravity and the center of pressure, so that the shaking of the person's body can be grasped.
According to the fifth aspect of the present invention, it is possible to determine whether or not a person has been hit by the longitudinal acceleration. Therefore, when analyzing the cause of a person's walking abnormalities, he / she is able to detect a person's physical condition and malfunction (for example, injury). Etc.) can be suppressed. Conversely, it is also possible to determine a person's physical condition and a physical defect (for example, an injury) from the situation where the whisper has occurred.

1 is a schematic block diagram of a posture and walking state estimation device 1 of the present embodiment. FIG. 4 is a schematic explanatory view showing an attachment state of the detection means 10. It is a schematic explanatory drawing of the state which attached to the test subject the apparatus used for experiment. (A) is the experimental result of (1), (B) is the experimental result of (2). (A) is explanatory drawing of the experimental method of (3), (B) is the experimental result of (3). (A) is the figure which compared the sensor coordinate system and the average walking stationary coordinate system, (B) is explanatory drawing of the method of calculating a pressure center.

  The posture and walking state estimation device 1 of this embodiment will be described with reference to the drawings.

  As shown in FIG. 1, the posture and walking state estimation device 1 of the present embodiment includes a detection unit 10 and an analysis unit 20. The detection means 10 detects a force applied to a human body and a displacement of the human body. The analysis unit 20 has a function of estimating the walking state and posture of a person based on the signal from the detection unit 10. The posture and walking state estimation device 1 of the present embodiment wears the posture and walking state estimation device 1 of the present embodiment by the analysis unit 20 analyzing the acceleration and speed detected by the detection unit 10. It is possible to estimate the walking state and posture of a person.

  The posture and walking state estimation device 1 of the present embodiment is used by being attached to a human body, but the analysis means 20 may be used by being attached to a human body or attached to a human body. You may use it in the state without.

  If the analysis means 20 is used without being attached to the human body, an advantage that the device attached to the human body can be reduced in size can be obtained. In this case, a storage device for storing the signal detected by the detection unit 10 may be provided, and the data stored in the storage device may be supplied to the analysis unit 20 later.

  On the other hand, when the analysis means 20 is used by being attached to a human body, the analysis result of the analysis means 20 is supplied to other equipment or the like (walking assist equipment or the like) worn by the person. It is possible to adjust the operating state of the device. Then, when a person walks with the assistance of other equipment or the like, it becomes possible to keep the person's walking constantly stable.

  Of course, if the communication function is provided in the detection means 10 or the analysis means 20, the signal from the detection means 10 can be supplied to the analysis means 20 even when the analysis means 20 is not attached to a human body. In addition, when the analysis result of the analysis unit 20 is supplied to another device, the communication unit of the analysis unit 20 can be provided even if the analysis unit 20 is not attached to a human body if the other device has a communication function. The analysis result can also be supplied to other devices.

(Detection means 10)
Details of the detection means 10 will be described.
The detection means 10 is used by being attached to a human trunk, and includes a rotation angle detection unit 11, a vertical acceleration detection unit 12, a lateral acceleration detection unit 13, and a longitudinal acceleration detection unit 14. I have.
Note that the attachment position of the detection means 10 is preferably attached in the vicinity of the belt position of the human trunk, and particularly in the vicinity of the spine at the belt position. If it is attached to such a position, the detection means 10 can be installed near the position of the center of gravity of the person, so that the advantage that the acceleration, the rotation angle, etc. can be measured with high accuracy can be obtained.

(rotation angle)
The rotation angle detector 11 is provided to detect the rotation of the human body, specifically, the rotation around the trunk. For example, a known angle sensor or the like can be used as the rotation angle detection unit 11. The position where the rotation angle detector 11 is attached to the human body is not particularly limited, but it is desirable that the rotation around the vertical axis is the first rotation as the Euler angle when a person walks. For example, the above condition can be satisfied if the trunk is attached to a position where the rotation around the vertical axis is small due to movements other than walking (for example, around the back of the belt). The reason why the rotation about the vertical axis is the first rotation as the Euler angle when the person walks is to accurately grasp the change in the body direction when the person walks.

(Vertical acceleration)
The vertical acceleration detector 12 is provided to detect the vertical acceleration of the human body. For example, a known acceleration sensor or the like can be used as the vertical acceleration detection unit 12. Although the position where the vertical acceleration detection unit 12 is attached to the human body is not particularly limited, when the human walks, the vertical movement of the human body can be measured with little disturbance (variation due to movement other than walking). It is desirable to install it at a position where it can. For example, the above condition can be satisfied by attaching the trunk near the belt position.

(Lateral acceleration)
The lateral acceleration detection unit 13 is provided to detect the acceleration in the horizontal direction of the human body. For example, a known acceleration sensor or the like can be used as the lateral acceleration detection unit 13. The position at which the lateral acceleration detector 13 is attached to the human body is not particularly limited, but when the person walks, the lateral movement of the human body is measured with little disturbance (variation due to movement other than walking). It is desirable to attach it at a position where it can. For example, the above condition can be satisfied by attaching the trunk near the belt position.

(Front-back acceleration)
The longitudinal acceleration detection unit 14 is provided to detect the lateral acceleration of the human body. For example, an acceleration sensor or the like can be used as the longitudinal acceleration detection unit 14. Although the position where the longitudinal acceleration detection unit 14 is attached to the human body is not limited, when the person walks, the movement of the human body in the front-rear direction is measured with little disturbance (variation due to movement other than walking). It is desirable to attach it at a position where it can. For example, the above condition can be satisfied by attaching the trunk near the belt position.

  If it is only necessary to grasp whether or not a person is walking and there is no need to analyze the shift of the center of gravity of the person or the movement of the left or right foot during walking, The direction acceleration detection unit 13 and the longitudinal acceleration detection unit are not necessarily provided.

  In addition, if the analysis unit 20 described later does not detect a whisper, the longitudinal acceleration detection unit 14 is not necessarily provided.

(Example of detection means 10)
In the above example, the case has been described in which the detection means 10 measures the rotation of the human body and the accelerations in the vertical and horizontal directions with separate sensors. However, a known posture sensor can be used as the detection means 10. A known posture sensor generally includes a three-axis accelerometer, a three-axis gyro (angular velocity measurement), a three-axis geomagnetic sensor (angular displacement measurement), etc. The rotation angle can be measured with one sensor. That is, all the functions of the rotation angle detection unit 11, the vertical acceleration detection unit 12, the lateral acceleration detection unit 13, and the longitudinal acceleration detection unit 14 can be exhibited with one posture sensor. In this case, in the posture sensor, the function of detecting the rotation around the trunk corresponds to the rotation angle detection unit 11, and the function of detecting the vertical acceleration of the human body corresponds to the vertical acceleration detection unit 12. The function of detecting the acceleration in the left-right direction of the body corresponds to the lateral acceleration detection unit 13, and the function of detecting the acceleration in the front-rear direction of the human body corresponds to the longitudinal acceleration detection unit 14.
In the following description, in the case of the rotation angle detection unit 11, the vertical acceleration detection unit 12, the lateral acceleration detection unit 13, and the longitudinal acceleration detection unit 14, the posture sensor performs the same function as each detection unit. It also includes the case to do.

(Calculation of acceleration of stationary coordinate system)
As the acceleration detected by each detection unit and posture sensor, the acceleration in the instantaneous stationary coordinate system (sensor coordinate system) of each detection unit and posture sensor is measured. Here, in the case of the above-described attitude sensor (that is, a sensor having a three-axis accelerometer, a three-axis gyro, and a three-axis geomagnetic sensor (hereinafter simply referred to as each sensor unit)), usually the acceleration and rotation angle measured by each sensor unit. The angular velocity and the Euler angle (attitude angle) of the attitude sensor are obtained as outputs. Therefore, if the Euler angles are used to coordinate the acceleration and angular velocity of each sensor unit, the acceleration in each axis direction and the angle and angular velocity around each axis direction in the stationary coordinate system can be obtained even when a person is walking. can get. That is, when the above-described attitude sensor is used as the detection means 10, the three-axis acceleration of the sensor's instantaneous stationary coordinate system (sensor coordinate system) is coordinate-converted into the stationary coordinate system, and each direction in the stationary coordinate system is converted. Acceleration can be calculated. Then, the acceleration in the vertical direction and other accelerations can be accurately calculated.

  For example, if the Euler angle obtained as the output of the attitude sensor is used to convert the triaxial acceleration of the sensor coordinate system (the output of the triaxial accelerometer) into the stationary coordinate system, the vertical direction obtained in the sensor coordinate system The acceleration in the vertical direction of the stationary coordinate system can be obtained from this acceleration.

In addition, as described above, after converting the three-axis acceleration of the sensor coordinate system to the stationary coordinate system, the angular displacement around the vertical axis of the stationary coordinate system is averaged for one period (two steps on the left and right) Find the average value. Then, the stationary coordinate system rotated by the average value obtained around the vertical axis from the initial stationary coordinate system is set. While being fixed to the set stationary coordinate system, the acceleration of the sensor coordinate system obtained during the one cycle is coordinate-converted to the set stationary coordinate system. Then, the acceleration in the walking direction and the acceleration in the direction orthogonal to the walking direction (that is, the horizontal direction) (that is, the acceleration in two directions orthogonal to the horizontal) are obtained.
When it is desired to grasp the walking direction in real time, the average value of the angular displacement around the vertical axis of the stationary coordinate system is obtained for the previous step or several steps before that instead of the average value for one cycle. If the average value is used, the acceleration in the walking direction and the acceleration in the direction orthogonal to the walking direction can be obtained by the same method as described above.

The detecting means 10 itself may have a function of calculating the acceleration of the stationary coordinate system by converting the three-axis acceleration (output of the three-axis accelerometer) of the sensor coordinate system, as described above, The analysis means 20 may have and is not specifically limited. In addition, when the detection means 10 has a coordinate conversion function, the detection means 10 will have a well-known attitude | position sensor and a coordinate conversion part.
In the following description, a case will be described in which the analysis unit 20 estimates the walking state using the acceleration or the like coordinate-converted by the detection unit 10 in principle.

(Analyzing means 20)
Details of the analysis means 20 will be described.
The analysis unit 20 analyzes the walking state of a person (hereinafter referred to as a subject) to which the detection unit 10 is attached based on a signal from the detection unit 10. The analysis means 20 includes a walking determination function 21, a posture stability determination function 22, and a whisper determination function 23.

(Walking judgment function 21)
The walking determination function 21 has a function of detecting whether or not the subject is walking, and in which direction, if the subject is walking, is detected. In brief, the walking determination function 21 determines whether or not a person is in a walking state based on a signal from the vertical acceleration detection unit 12, and walks on the person based on a signal from the rotation angle detection unit 11. It has a function to determine the direction.

(Walk detection)
The walking determination function 21 detects whether or not the vehicle is in a walking state by processing time fluctuations in the vertical acceleration detected by the vertical acceleration detection unit 12. Specifically, it has a function of estimating a vertical floor reaction force based on a signal from the vertical acceleration detection unit 12 and determining whether a person is in a walking state based on a temporal variation of the estimated floor reaction force. ing.

  The floor reaction force is usually obtained by measuring the force applied to the floor reaction force meter using a floor reaction force meter placed on the floor. For this reason, it was impossible to grasp the floor reaction force during walking.

  On the other hand, the floor reaction force has a floor reaction force in three directions, up, down, left, and right. It is considered that the floor reaction force in each direction is obtained by multiplying the acceleration in the three directions of the center of gravity by the mass. For this reason, it is considered that the vertical floor reaction force described above is obtained by multiplying the acceleration in the vertical direction of the center of gravity by the mass. Therefore, in the posture and walking state estimation device 1 of the present embodiment, the walking determination function 21 uses the three-way acceleration of the center of gravity detected by the detection means 10 to temporally vary the three-way floor reaction force including the vertical floor reaction force. Is estimated.

  Here, the vertical floor reaction force increases when an upward acceleration occurs, and the vertical floor reaction force decreases when a downward acceleration occurs. When a person is walking with heel contact, the vertical floor reaction force generally varies with time so that one foot has two upward peaks in the period from heel contact to toe separation. To do. The first peak occurs immediately after the heel contact, and the second peak occurs at the timing when the foot opposite to the heel contact is separated from the toe.

  Therefore, the walking determination function 21 determines the walking state based on the fluctuation of the vertical acceleration, and pays attention to the first peak that occurs immediately after the heel-contact, and the free leg period (the foot is separated from the ground). It is determined whether or not the feet are in contact with the ground.

  Even when a person is stepping on, the person raises and lowers the left and right feet alternately, and thus vertical acceleration occurs. However, when walking, acceleration / deceleration is repeated in the walking direction (front-rear direction). Therefore, if a change in acceleration in the walking direction is confirmed, stepping and walking can be distinguished.

(Judgment of left and right feet)
Moreover, it is desirable that the walking determination function 21 also has a function of detecting which of the left and right feet is grounded. A healthy person usually walks by alternately stepping on the left and right feet, so that having such a function makes it possible to grasp a person's walking more accurately.

For example, the walking determination function 21 detects which foot is grounded by processing the time variation of the floor reaction force and the time variation of the lateral acceleration detected by the lateral acceleration detection unit 13. Can do.
First, when a person is walking with the left and right feet alternately raised and lowered, vertical acceleration occurs in the human body according to the movement of the feet. The upward acceleration peaks at the timing.
On the other hand, when walking, the lateral acceleration also fluctuates with time, and rightward acceleration occurs during the period when the left foot is grounded, and leftward acceleration occurs during the period when the right foot is grounded.
Therefore, when the acceleration in the vertical direction is the maximum, it is possible to detect which of the left and right feet is in contact with the ground depending on which direction the acceleration is generated in the left or right direction.

  Note that it may be determined which foot is in contact with the displacement of the center of gravity in the left-right direction. In human walking, the center of gravity is from the left when the left foot is grounded, and the center of gravity is from the right when the right foot is grounded. Therefore, when the acceleration in the vertical direction is maximized, it can be detected whether the left or right foot is in contact with the ground by determining whether the center of gravity is biased in the left or right direction.

  The method for grasping the displacement of the center of gravity is not particularly limited. For example, the displacement can be obtained by integrating the acceleration in the left-right direction twice. However, in this case, the obtained displacement amount may drift and become larger than the actual displacement. As a method of preventing such drift, in the case of batch processing, when calculating the speed from the acceleration, it is calculated for one period, corrected so that the average value becomes zero, and integrated to obtain the displacement. The effect of drift can be eliminated.

(Walking direction detection)
The walking determination function 21 has a function of detecting a direction in which a person is moving by processing a time variation of rotation around a vertical axis (Z axis) detected by the rotation angle detection unit 11. Have. Specifically, the direction rotated by the average rotation angle around the vertical axis in one cycle (2 steps) or half cycle (1 step) by analyzing the signal from the rotation angle detector 11 Can be the walking direction. In this case, if the rotation angle detector 11 is attached at a position where there is little rotation around the vertical axis due to other movements, fluctuations in rotation around the vertical axis can be accurately detected. It is possible to accurately grasp the walking direction.

As described above, it is desirable to determine the walking direction based on the angular displacement around the vertical axis of the stationary coordinate system by converting the acceleration obtained in the sensor coordinate system into the stationary coordinate system.
In addition, when this method is adopted, it is possible to grasp a blurring case of a walking person in the walking direction.

As for the angular velocity around the vertical axis, if the average value of one cycle (or half cycle) of walking is obtained, the walking direction can be estimated using the angular velocity. For example, estimates of the average angle of the next two steps or one step, if a period between t _i, and a mean angle of the immediately preceding average angular velocity, can be determined by the following equation (1).
Estimated average angle:
Previous average angle:
Last average angular velocity:

That is, the next moving direction can be estimated from the previous walking state.

  The method for determining the walking direction is not limited to the above method, and various methods can be adopted.

(Estimation of walking speed)
The walking determination function 21 may have a function of estimating the walking speed. Then, a person's walking state can be grasped in detail. As a method of estimating the walking speed, the following method is used, that is, a method of obtaining the average walking speed by integrating the acceleration in the walking direction in the walking direction stationary coordinates (corresponding to the average walking stationary coordinate system described later). Can be adopted.

  First, walking is started in a stationary state, and after walking to some extent, it stops and ends. At this time, the number of steps and the time to walk are not particularly limited, but are preferably within 10 steps.

Next, the walking speed is estimated from the measurement result by the following method.
When the time from the start of measurement to the end of measurement is T, the initial speed is set to zero, the acceleration is integrated for T time, and the speed at the end is calculated. Since the speed should be zero at the end, the final speed obtained by integration is divided by T to obtain an acceleration α, and this acceleration α is subtracted from the measured acceleration (see Equation 2). Then, if the acceleration obtained by subtraction is integrated, the final speed becomes zero.
Final speed:
Measured acceleration:

  Then, the average walking speed can be obtained by obtaining the average speed during the period when the speed is stable at the speed obtained so that the final speed becomes zero. If such a method is used, it is possible to prevent a decrease in accuracy of the calculated average speed due to a drift at the time of integration.

(Attitude stability judgment function 22)
The analysis means 20 has a posture stability determination function 22 for determining whether the posture of the person is stable and whether the person is walking stably. If it has such a function, it is possible to grasp abnormal changes in the posture of the person and the walking state of the person. ) May be possible.

  The posture stability determination function 22 grasps the time variation of the center of gravity position based on the relative position between the detection unit 10 and the center of gravity of the human body (input in advance) and the signal from the detection unit 10. The stability of the walking state of the person is determined based on the temporal variation of the center of gravity position. The center-of-gravity position can be obtained from the center-of-gravity position and the inclination angle in the human coordinate system of the attachment position of the detection means 10 and the distance between them.

  Here, if the time variation of the centroid position, that is, the acceleration at the centroid position is not required with a very high accuracy, the acceleration at the centroid may be the same as the position where the detection means 10 is attached. However, the center of gravity of the human body is generally located inside the detection means 10 and the accelerations of the two do not match. Therefore, when obtaining the acceleration of the center of gravity from the acceleration detected by the detection means 10, it is desirable to correct using the angular velocities around the three axes detected by the detection means 10 to obtain the acceleration of the center of gravity.

  On the other hand, if it is normal walking, the rotational speed other than around the vertical axis is considered to be small, so if only the angular velocity around the vertical axis is used and corrected by the following method, the acceleration of the center of gravity can be obtained with a certain degree of accuracy. Can do.

  First, there are three types of stationary coordinate systems that are set when a person walks. Absolute stationary coordinate system (initial coordinate system, stationary coordinate system described above), average walking stationary coordinate system (stationary coordinate system using average walking direction), instantaneous stationary coordinate system of detecting means 10 (stationary in the instantaneous posture of the sensor) Coordinate system, sensor coordinate system).

  The measurement value of the detection means 10 is a measurement value in the sensor coordinate system. In this coordinate system, the relationship between the attachment position of the detection means 10 and the position of the center of gravity does not change even when a person walks.

As shown in FIG. 6A, the acceleration of the center of gravity in the sensor coordinate system can be expressed by the following equation (3). If this acceleration is transformed into the average walking stationary coordinate system, the acceleration of the center of gravity in the walking direction and the lateral direction can be obtained.
The angular acceleration of the center of gravity around each axis in the average walking stationary coordinate system is approximately the angular velocity around each axis of the average walking stationary coordinate system obtained by converting the acceleration of Equation 3 into the average walking stationary coordinate system. Is obtained by differentiating.

The above method can be easily expanded when correcting the acceleration of the center of gravity using the biaxial angular velocity. That is, when the angular velocity of two axes (for example, the vertical axis and the X axis or the vertical axis and the Y axis) is used, the straight line connecting the detection means 10 and the center of gravity is set as the Y axis of the instantaneous stationary coordinate system of the detection means 10. Then, the acceleration of the center of gravity in the sensor coordinate system can be expressed by the following formula 4. If this acceleration is coordinate-converted into the average walking stationary coordinate system, the acceleration in the walking direction and the lateral direction of the center of gravity can be obtained even if the biaxial angular velocity is used.

  The posture of the person and the walking state of the person can be determined based on the temporal variation of the relative positional relationship between the position of the pressure center and the position of the center of gravity. Specifically, assuming that inertial forces in three directions (including gravitational acceleration in the vertical direction) are applied to the center of gravity of the human body, the coordinates of the pressure center viewed from this center of gravity are obtained. Then, by measuring the temporal variation of the relative positional relationship between the position of the pressure center and the position of the center of gravity, it is possible to determine the stability of the person's posture and the walking state of the person.

  For example, for people of various body types, people of various ages, and people with motor dysfunction (eg hemiplegia), the relative positional relationship between the position of the center of pressure and the position of the center of gravity when walking Time variation (hereinafter referred to as relative variation) is measured. And in each group, the pattern of the relative fluctuation in the state where walking is stable and / or the state where walking is unstable is put together as data. Then, by comparing the relative variation measured when the subject walks with the pattern of the relative variation of the group to which the subject belongs, it can be determined whether or not the subject's walking is stable.

(Calculation method of inertial force applied to the center of gravity)
The vertical inertia force applied to the position of the center of gravity described above can be obtained from the product of the equivalent mass and the vertical acceleration in accordance with the walking phase determined from the timing of heel contact. That is, the vertical floor reaction force applied to the human foot is estimated as the inertial force.

Here, as for the equivalent mass, as the simplest method, the mass of the subject can be adopted as the equivalent mass.
On the other hand, in order to accurately grasp the equivalent mass, ask the subject to walk on the floor reaction force and simultaneously measure the vertical acceleration. Then, the fluctuation | variation of the equivalent mass accompanying a test subject's walk (namely, the equivalent mass for every walk phase) can be grasped | ascertained. In this way, if the equivalent mass for each walking phase of the subject is calculated in advance, it is possible to accurately grasp the fluctuation of the vertical inertia force applied to the position of the center of gravity accompanying the walking of the subject.

As an intermediate method between the above two methods, the equivalent mass can be calculated by the following method.
First, the equivalent mass in the vertical direction can be expressed as the following equation (5), assuming that the substantial mass is m and the conversion count is η _z when both feet are in contact with the ground.

The conversion coefficient η _{z at} the moment of contact with the heel is defined as follows.
η _z = η _z0

If the period of one step is T, the conversion coefficient η _z is defined as follows during the period 0 ≦ t ≦ αT.
η _z = η _z0 + (1−η _z0 ) t / α

Further, when αT ≦ t ≦ αT, the conversion coefficient η _z is defined as follows.
η _z = 1

Note that η _z and α may use the average value of a large number of subjects (obtained from accurate simultaneous measurement of floor reaction force), but if there are measured values of the subjects, the individual data of the subjects can be used. desirable.

Further, the inertial force applied in the left-right front-rear direction can be obtained by the following method.
For example, the acceleration at the detection means mounting position and the acceleration at the center of gravity can be the same, and an inertial force applied in the left-right front-rear direction can be obtained by multiplying the acceleration by an equivalent mass. If this method is used, the inertial force applied in the left-right and front-back directions can be easily obtained.

  In addition, the magnitude of the inertial force in each direction applied to the center of gravity can be considered to coincide with the magnitude of the floor reaction force in each direction (inertia force = −floor reaction force). Therefore, as described in paragraph 0032, the floor reaction force in each direction (vertical floor reaction force, front-rear direction floor reaction force, left-right floor reaction force, left and right) is calculated from the acceleration in each direction and the equivalent mass by the above-described method (acceleration is multiplied by the equivalent mass). Directional floor reaction force).

  On the other hand, if the acceleration of the center of gravity is calculated using the method described in paragraphs [0050] to [0055] and the acceleration is multiplied by the mass as in the above method, the inertial force applied in the left-right and front-back directions can be obtained with higher accuracy.

(Calculation method of pressure center)
The pressure center described above can be obtained by the following method.
In FIG. 6B, thin arrows indicate acceleration applied to the person, and thick arrows indicate inertial force applied to the person. In FIG. 6B, the large circle indicates the center of gravity of the person, and the small circle indicates the center of pressure. G indicates the ground.

  First, the detection means 10 detects gravitational acceleration as a force applied to a person. That is, the detection means 10 detects the gravitational acceleration as an upward acceleration. Therefore, the vertical acceleration applied to the person is the sum of the acceleration of gravity applied to the person and the acceleration of movement of the person.

Here, the moment around N in FIG. 6B is zero. Therefore, in FIG. 6B, the following formula 6 is established.

When the above equation 1 is modified, therefore, the distance Δx from the attachment position of the detection means 10 or the center of gravity to the center of pressure in the walking direction is obtained by the following equation 7.

  Also in the horizontal direction, the distance pressure center from the attachment position of the detection means 10 or the center of gravity to the center of pressure is obtained in the same procedure.

Accordingly, the distance from the attachment position of the detection means 10 or the center of gravity to the center of pressure is obtained.

  If the distance from the center of gravity to the center of pressure can be grasped, it becomes possible to determine the stability of the posture of the person as follows. For example, if the person is standing without standing (still standing), the distance from the center of gravity to the center of pressure changes if the person fluctuates. Therefore, if a change in the distance from the center of gravity to the center of pressure is detected, it is possible to detect blurring of the human body. That is, the posture and walking state estimation device 1 of the present embodiment can also be used as a sway meter. Specifically, the blur of the pressure center at rest can be obtained by the following method.

Since the average value of the longitudinal and lateral accelerations can be regarded as zero when the vehicle is stationary, the acceleration of the center of gravity in the stationary coordinate system that has been measured and coordinate-transformed is integrated to obtain the fluctuation displacement _xGA . If the obtained variable displacement x _GA is added to Δx, it is possible to obtain an output (Δx + x _GA ) equivalent to a so-called center-of-gravity sway meter. Then, even if one motion sensor is used as the detection means 10, it becomes possible to make the center of gravity oscillating meter work.

Therefore, the fluctuation of the center of gravity can be obtained by the variable displacement x _GA , and the fluctuation of the pressure center can be obtained by Δx + x _GA . The relationship between the center of gravity and the center of pressure can be obtained from Δx.

  And even when walking, if the above calculation is performed for each cycle (period of walking two steps left and right), the sway of the center of gravity, the sway of the pressure center, and the relationship between the center of gravity and the center of pressure can be obtained. Can do.

Incidentally, the pressure center during walking, may be the aforementioned [Delta] x _G. Further, Δx _G may be the value obtained by adding the above-described displacement fluctuation value Δx. In this case, the pressure center changes with respect to the average speed during walking. Each index has meaning, and can be appropriately selected and used.

Note that the integration error in determining the fluctuation displacement x _GA by integrating the acceleration of the center of gravity, since there is no variation in one period as walking, cancellation by a zero mean value of the velocity in the relatively long time can do.
In addition, as a method of setting the average value to zero, the measured acceleration may be covered with a window, Fourier transformed to remove the direct current (DC) component, and the frequency component multiplied by 1 / ω to obtain the velocity by inverse Fourier transformation. Good. Furthermore, the displacement can also be obtained by multiplying the frequency component by 1 / ω ² and Fourier transform.

(Estimation of stride)
If the center of pressure during walking can be grasped by the above-described method or the like, it becomes possible to estimate the stride of a person.

  When a person is walking and only one foot is in contact with the ground, the center of pressure is located in the sole. For this reason, if the relative fluctuation range of the distance between the position of the pressure center and the position of the center of gravity or the attachment position of the detection means 10 is grasped, a rough stride can be grasped.

  Specifically, in a state where one foot is in contact with one foot, the position of the center of gravity moves from the rear to the front of the center of pressure in the direction of human movement. The distance from the center of gravity position to the center of pressure is the farthest in the state where both feet are grounded and one foot is grounded. Then, if the forward direction in the direction of travel is positive, the state from the center of gravity to the center of pressure is such that the state from the ground contact to the one foot contact is the negative maximum value and the state from the one foot contact to the both foot contact is the positive maximum value The distance changes.

  Therefore, if the fluctuation range of the pressure center is grasped, the distance between the position of the pressure center in the state where the ground contact from one foot is in contact with one foot and the position of the pressure center in the state where the ground contact from both feet is contacted, that is, the stride is roughly grasped. It can be done.

(Whiggle judgment function 23)
The analysis means 20 preferably has a whisper determination function 23 for determining whether or not a person has struck. If it has such a function, it can be judged whether it is caused by a change in the walking state of a person or for another reason. Then, when judging the physical condition of the subject or the malfunction of the subject (for example, an injury) from the abnormality in the walking of the subject, it is possible to prevent mistaking the whisper as another state. Conversely, it is also possible to determine a person's physical condition and a physical defect (for example, an injury) from the situation where the whisper has occurred.

  The following method can be cited as a method for the whisper determination function 23 to determine whisper.

  First, when a person is walking in a stable state, the longitudinal acceleration detected by the longitudinal acceleration detector 14 is a steady-state vibration waveform that fluctuates around zero. And the width of the steady vibration is within a predetermined range.

  However, when the fluctuation of the longitudinal acceleration detected by the longitudinal acceleration detection unit 14 exceeds a predetermined range, it indicates that the posture of the person is greatly collapsed. That is, the change in the posture of the person can be grasped to some extent by the amplitude of the acceleration in the front-rear direction.

  Here, whispering is a state in which the foot is grounded quickly without being able to move the foot forward at the normal walking timing due to something such as a foot that is not grounded being caught. Therefore, it can be determined that an abnormality such as a whisper has occurred if the acceleration in the vertical direction increases upward almost simultaneously with the amplitude of the acceleration in the front-rear direction exceeding a predetermined range. Then, by referring to the lateral acceleration or displacement when an abnormality such as whispering is detected, it is possible to grasp which foot has been hit.

  Then, by comparing the timing of grounding in a normal walking state with the timing of installation at the time of hitting, it is possible to determine at which point the user hits. For example, if the ground contact timing (the timing at which the acceleration in the vertical direction increases upward) is earlier than the ground contact timing for walking until the detection of a whisper, It can be determined that he / she crawls during the floating period (free leg period).

  While the above method can determine the occurrence of whispering, it is also possible to estimate that walking has been disturbed for reasons other than whispering. For example, even if the amplitude of the acceleration in the front-rear direction exceeds a predetermined range, it can be estimated that the walking is disturbed due to a cause other than whispering if a phenomenon in which the acceleration increases upward in the vertical direction does not appear.

  Moreover, the method by which the whispering determination function 23 determines whispering is not limited to the method as described above. When the amplitude of acceleration in the front-rear direction exceeds a predetermined range and the rate of change (acceleration) of the rotational angular velocity detected by the rotational angle detection unit 11 increases, it may be determined that a whisper has occurred. . For example, if the angular acceleration around the horizontal axis (Y axis) in the sensor coordinate system is large, it indicates that the vehicle is falling forward. Therefore, even when the amplitude of the acceleration in the front-rear direction exceeds a predetermined range and the angular acceleration around the horizontal axis (Y-axis) in the sensor coordinate system becomes large, the whirling occurs and it tends to fall forward Can be determined.

  If the angular acceleration around the axis in the sensor coordinate system is used, it is possible to estimate the direction of falling when hit. For example, when the angular acceleration around the walking direction axis (X axis) is larger than normal, it can be determined that the direction of falling is not purely forward but is tilted diagonally. Of course, even when the acceleration in the lateral direction (Y-axis direction) is larger than usual, it can also be determined that the direction of falling is not purely forward but is tilting diagonally.

  Furthermore, when determining the stroke, the stroke may be determined only by the change rate (acceleration) of the rotational angular velocity without using the amplitude of the acceleration in the front-rear direction. That is, when the change rate (acceleration) of the rotation angular velocity detected by the rotation angle detection unit 11 is in the state as described above, it may be determined that a whisper has occurred. However, the accuracy of detecting whispering can be increased by using the acceleration amplitude in the front-rear direction.

  It was confirmed that the walking state can be estimated by the posture and walking state estimation device of the present invention.

  In the experiment, with the posture and walking state estimation device of the present invention, (1) whether or not the foot is grounded, (2) whether the left or right foot is grounded, or (3) It was confirmed that it was possible to detect whether there was a possibility of falling due to a whisper.

  In the experiment, as shown in FIG. 3, the walking state was estimated based on information from a motion sensor (MSD-MS manufactured by Tech Gifu Co., Ltd.) that is attached to the trunk and requires acceleration in three axial directions. In the motion sensor, data was measured at a sampling period of 500 Hz while a person was walking, and the obtained data was analyzed to estimate the states (1) to (3).

  In order to confirm the validity of the estimation results, a walking floor reaction force meter (M3D-FP: Tao Liu et al, “A Mobile Force Plate and Three-Dimensional Motion, manufactured by Tech Gifu) Wearing the “Analysis System for Three-Dimensional Gait Assessment”, IEEE SENSORS JOURNAL, VOL. 12, NO. 5, MAY 2012), the reaction force applied to the foot contact was measured.

(1) Whether or not the foot is in contact with the ground FIG. 4A shows the result of confirming whether or not the foot installation can be grasped by the posture and walking state estimation device of the present invention. In Fig. 4 (A), RHTSRIO and LHTSRIO indicate on / off of ground contact between the left and right legs in the floor reaction force meter. When the value is 5, the foot is grounded to the ground, and when it is 0 Shows that your feet are floating.

  As shown in FIG. 4A, it was confirmed that the acceleration in the Z-axis direction of the trunk sensor was increased when it was considered that the foot was in contact with the ground. That is, it was confirmed that it is possible to grasp whether or not the foot is in contact with the ground by detecting the acceleration in the Z-axis direction (that is, the vertical direction).

(2) Which of the left and right feet is grounded? FIG. 4 (B) shows the result of confirming whether the right and left feet can be grounded by the posture and walking state estimation device of the present invention. FIG. 4B shows the timing of ground contact estimated from the dynamic acceleration in the Z-axis direction and the value of the dynamic acceleration in the Y-axis direction. When the ON / OFF value is 5, it indicates the timing at which the foot touches the ground.

As shown in FIG. 4B, at the timing of grounding estimated from the dynamic acceleration in the Z-axis direction, the dynamic acceleration in the Y-axis direction is leftward when the right foot is grounded and rightward when the left foot is grounded. I was able to confirm. In other words, it was confirmed that by comparing the direction of the dynamic acceleration in the Y-axis direction and the acceleration in the Z-axis direction, it is possible to grasp which of the left and right feet is grounded.
Here, the X axis and the Y axis are an average walking stationary coordinate system. In this experiment, since the subject is walking straight, the average walking stationary coordinate system matches the absolute stationary coordinate system. That is, the Z axis coincides with the vertical axis, and the X axis and the Y axis are horizontal axes.

(3) Is there a possibility of a fall due to a roll? It was confirmed whether a roll can be grasped by the posture and walking state estimation device of the present invention.
In the experiment, as shown in FIG. 5 (A), the whip generated during walking was reproduced by pulling a string attached to the foot.

  In the experiment, five subjects performed normal walking three times and left and right whispers three times. The experiment was conducted in a situation where all subjects were fully explained about the purpose and method of the experiment, agreed to cooperate, and sufficiently secured.

  FIG. 5B shows an example of the difference between the change in the dynamic acceleration in the X-axis direction seen when whispering and the normal walking. As shown in FIG. 5 (B), it can be confirmed that the dynamic acceleration in the X-axis direction greatly increases or decreases when whispering compared to when walking normally.

  Table 1 shows the range of dynamic acceleration obtained from the average of the maximum and minimum values for each subject. In Table 1, it was confirmed that every subject increased the increase / decrease width of the dynamic acceleration when stroking as compared to the normal walking.

  From the above results, it was confirmed that the whispering can be detected by detecting the dynamic acceleration in the X-axis direction.

  The posture and walking state estimation device of the present invention is a device that is attached to a human body and is used to determine a physical condition of the subject or a malfunction of the body (for example, an injury) from a human walking abnormality, or a device that diagnoses a fall prevention ability It can be applied to a device for quantifying results of walking training and the like.

DESCRIPTION OF SYMBOLS 1 Posture and walking state estimation apparatus 10 Detection means 11 Rotation angle sensor 12 Vertical acceleration sensor 13 Lateral acceleration sensor 14 Longitudinal acceleration sensor 20 Analysis means 21 Walking state judgment function 22 Posture stability judgment function 23 Whisper judgment function

Claims (5)

Detection means attached to the human body;
Analyzing means for analyzing a walking state and / or posture based on a signal from the detection means,
The detection means is
It has a vertical acceleration detector that detects the vertical acceleration of the human body,
The analysis means includes
It has a posture stability judgment function to judge the posture stability,
The posture stability determination function is
A function of estimating a vertical floor reaction force based on a signal from the vertical acceleration detection unit and determining a walking state and / or posture based on a temporal variation of the estimated vertical floor reaction force; Characteristic posture and walking state estimation device. The posture stability determination function is
Based on the relative position between the detection means and the gravity center position of the human body and the signal from the detection means, the change in the gravity center position is grasped, and the person walking based on the time variation of the gravity center position 2. The posture and walking state estimation device according to claim 1, wherein the posture and walking state stability are determined. The analysis means includes
The inertial force applied to the center of gravity position of the human body is estimated, and the stability of the walking state and / or posture of the person is determined based on the temporal variation of the center of gravity position in the state where the inertial force is applied. The posture and walking state estimation device according to claim 1 or 2. The posture stability determination function is
4. The posture according to claim 1, 2, or 3, wherein the posture of the person and / or the stability of the posture are determined based on a relative positional relationship between the center of gravity position and the center of pressure. Walking state estimation device. The detection means includes
It has a longitudinal acceleration detector that detects the longitudinal acceleration of the human body,
The analysis means includes
The posture and walking state estimation according to claim 1, 2, 3, or 4, further comprising a function of determining a whirling based on a signal from the longitudinal acceleration detection unit and a temporal variation of the floor reaction force. apparatus.