JP5200920B2 - Walking assist device - Google Patents

Description

  The present invention relates to a walking assist device that assists a user's walking by applying torque to a joint of a leg. In particular, the present invention relates to a walking assistance device that assists a user who cannot freely move one leg in the same manner as the other leg. Hereinafter, a leg that can be moved freely is referred to as a healthy leg, and a leg that cannot be moved freely is referred to as an affected leg.

  A walking assist device that assists the user's walking by applying torque to the joint of the affected leg has been studied. For example, Patent Document 1 discloses a walking assist device that applies torque to a joint of an affected leg so as to be the same as the movement of a healthy leg.

JP 2006-314670 A

  A user targeted by the walking assist device of Patent Document 1 cannot move the affected leg in the same manner as a healthy leg. That is, the trajectory drawn by the healthy leg is different from the trajectory drawn by the affected leg. Therefore, the walking assist device that applies torque to the joint of the affected leg so as to be the same as the movement of the healthy leg forces the user to be forced. On the other hand, an apparatus that assists the joint of the affected leg so as to follow only the target trajectory of the affected leg in advance according to the user cannot perform appropriate assistance. Since walking is a movement in which the movements of both legs are synchronized, assisting the affected leg ignoring the movement of the healthy leg tends to cause a deviation from the movement of the healthy leg. There is a demand for a walking assist device that smoothly supports movement of an affected leg while allowing the affected leg to be unable to move as well as a healthy leg.

  If the user's walking motion in which one leg cannot be freely moved is closely observed, the stride when stepping on the affected leg is smaller than the stride when stepping on the healthy leg. Here, the “step length” means the distance between the two toes when one leg is stepped on. In this specification, the stride when the healthy leg is stepped on is referred to as “health leg stride”, and the stride when the affected leg is stepped on is referred to as “affected leg stride”. Also, the time required to step out the affected leg is longer than the time required to step out the healthy leg. In the present specification, the ratio of the stride length of the affected leg to the stride length of the healthy leg is referred to as a stride ratio, and the ratio of the time required to step out the affected leg with respect to the time required to step out the healthy leg is referred to as a step time ratio. In general, the stride ratio is less than 1 and the step time ratio is greater than 1. The stride ratio and the step time ratio are substantially constant during walking. The inventors have found that the relationship between the trajectory of the healthy leg and the trajectory of the affected leg during walking correlates with the step ratio and the step time ratio.

  The present invention was created based on this finding. The technical idea of the present invention is as follows. An ideal walking pattern in which the curves drawn by the trajectories of both legs are substantially the same is prepared in advance, and the trajectory of one leg in the walking pattern is expanded or contracted by the step ratio, the step time ratio, or both. Torque is applied to the joint of the affected leg so as to follow the stretched walking pattern while synchronizing with the movement of the healthy leg. By providing such a mechanism, it is possible to assist the movement of the affected leg while allowing the affected leg to draw a trajectory different from the trajectory of the healthy leg.

  Hereinafter, the affected leg of the user is referred to as “first leg”, and the healthy leg is referred to as “second leg”. This is because it is not necessary to distinguish between a healthy leg and an affected leg in characterizing the technology of the walking assist device.

The walking assist device of the present invention includes a storage device, a leg sensor, an actuator, and a controller. The storage device stores a reference walking pattern. The reference gait pattern is time-series data describing the trajectory of the legs during walking in which the trajectories of both legs draw the same curve. That is, the reference walking pattern represents an ideal trajectory when both legs can be moved equally. The leg sensor detects the movement of each leg of the user. The actuator is attached to the user and applies torque to the joint of the first leg. The controller controls the actuator in synchronization with the movement of the user's second leg detected by the leg sensor. Here, the controller executes the following processing.
(1) A process of generating a target free leg trajectory in which the trajectory of the first leg in the reference walking pattern is expanded or contracted by a factor other than 1 along either the time axis or the longitudinal distance axis. Note that the trajectory of the first leg in the standard walking pattern may be expanded and contracted on both the time axis and the longitudinal distance axis. Further, in the reference walking pattern, the trajectory at the time of the free leg of the first leg is equal to the trajectory at the time of the free leg of the second leg. The purpose of “the trajectory of the first leg in the standard walking pattern” is to make the motion timing in the target leg trajectory generated based on the standard walking pattern correspond to the motion timing of the first leg of the user. It is necessary.
(2) A process of controlling the actuator so that the movement of the first leg during the free leg follows the target free leg trajectory.

  This walking assistance device generates a target trajectory for the affected leg from the leg trajectory for normal legs (reference walking data) by introducing “a coefficient other than 1”. The “non-1” coefficient may be the above-described stride ratio or step time ratio. These ratios may be obtained in advance by measurement or the like, or may be obtained in real time from the walking motion of the user wearing the walking assistance device. In any case, the walking assist device of the present invention differs from the normal leg trajectory by expanding and contracting the trajectory of the first leg in the reference walking pattern in which the trajectory curves of both legs are the same by the coefficient. Torque can be applied to the joint of the affected leg so as to follow the trajectory. The “non-one coefficient” is a coefficient that determines the asymmetry between the trajectory of the healthy leg and the trajectory of the affected leg, and is hereinafter referred to as an asymmetry coefficient.

When the step ratio as the asymmetry coefficient is obtained in real time, the process (1) is as follows.
(1-a1) A process of determining a step ratio between the step length of the first leg and the step length of the second leg as an asymmetry coefficient based on the output of the leg sensor.
(1-a2) A process of generating a target free leg trajectory in which the trajectory of the first leg in the reference walking pattern is expanded and contracted by the asymmetry coefficient along the distance axis in the front-rear direction of the leg.

When the step time ratio as the asymmetry coefficient is obtained in real time, the process (1) is as follows.
(1-b1) A process of determining, as an asymmetry coefficient, a time ratio between the time required to step one step on the first leg and the time required to step one step on the second leg based on the output of the leg sensor.
(1-b1) Processing for generating a target free leg trajectory in which the trajectory at the time of the free leg of the first leg in the reference walking pattern is expanded and contracted along the time axis by an asymmetry coefficient.

  It is preferable that the controller asymptotically approaches the asymmetry coefficient to 1 as the use time of the walking assist device elapses. When the step ratio and the step time ratio are obtained in real time, the asymmetry coefficient may gradually approach 1 as the user's walking ability recovers, or the walking assist device positively sets the asymmetry coefficient to 1. It may be asymptotic. An asymmetry factor of 1 means that the target free leg trajectory is the same as the trajectory of the free leg in the reference walking pattern. That is, when the asymmetry coefficient is asymptotic to 1, the asymmetry between the free leg trajectory and the target free leg trajectory in the reference walking pattern gradually decreases. By making the asymmetry coefficient asymptotically asymptotic to 1, recovery of the user's walking ability can be promoted.

  ADVANTAGE OF THE INVENTION According to this invention, the walking assistance apparatus which assists a motion of an affected leg smoothly can be implement | achieved, accept | permitting that a user cannot move an affected leg equivalent to a healthy leg.

  (First Embodiment) A walking assistance device according to a first embodiment will be described with reference to the drawings. FIG. 1 schematically shows a walking assistance device 10 worn by a user. 1A shows a front view, and FIG. 1B shows a side view. In the present embodiment, it is assumed that the user cannot freely move the knee joint of the left leg. That is, the left leg corresponds to the affected leg, and the right leg corresponds to the healthy leg. The left leg corresponds to the first leg described in the claims, and the right leg corresponds to the second leg described in the claims. The walking assist device 10 assists the user's walking motion by applying appropriate torque to the knee joint of the left leg (affected leg, first leg) while synchronizing with the movement of the right leg (normal leg, second leg). .

  The walking assist device 10 includes a right leg orthosis 12R, a left leg orthosis 12L, and a support bar 28 connecting them. The support bar 28 is disposed on the back side of the user and connects the upper end of the right leg orthosis 12R and the upper end of the left leg orthosis 12L.

  The right leg orthosis 12R is attached to the outside of the right leg from the user's thigh to the lower leg. The right leg orthosis 12R has an upper link 14R and a lower link 16R, and the two links are rotatably connected. The upper link 14R is fixed to the user's thigh with a belt. The lower link 16R is fixed to the user's lower leg with a belt. The connecting portion between the upper link 14R and the lower link 16R is located outside the right knee joint. The connecting portion is provided with a knee encoder 20R and detects the angle of the right knee joint.

  The upper end of the upper link 14R is connected to the support bar 28 via a rotary joint. A waist encoder 22R is attached to the rotary joint. The waist encoder 22R detects an angle around the pitch axis of the right hip joint.

  A ground sensor 24R is attached to the lower end of the lower link 16R. The ground sensor 24R detects the right leg leaving timing and the ground timing.

  The left leg orthosis 12L is attached to the outside of the left leg from the user's thigh to the lower leg. The left leg orthosis 12L has the same structure as the right leg orthosis 12R. That is, the left leg orthosis 12L includes an upper link 14L and a lower link 16L, and the two links are connected to be rotatable outside the left knee joint. The upper end of the upper link 14L is connected to the support bar 28 via a rotary joint. The left leg orthosis 12L includes a knee encoder 20L, a waist encoder 22L, and a grounding sensor 24L. The knee encoder 20L detects the angle of the left knee joint, and the waist encoder 22L detects the angle of the left hip joint. The knee joint angle and the hip joint angle mean a rotation angle around a pitch axis (an axis extending in the body side direction). The ground sensor 24L detects the left leg leaving timing and the ground timing.

  The left leg orthosis 12L also includes a motor 26. The motor 26 is provided at a connecting portion between the upper link 14L and the lower link 16L, and is located outside the user's knee joint. The motor 26 can rotate the lower link 16L with respect to the upper link 14L. That is, the motor 26 can apply torque to the user's left knee joint. The left leg orthosis 12L corresponds to the actuator described in the claims.

  A controller 30 is attached to the support bar 28. A tilt sensor 27 is provided inside the controller 30. The tilt sensor 27 detects an absolute tilt angle of the user body. That is, the tilt sensor 27 detects the tilt angle of the user body with respect to the vertical direction. The controller 30 controls the motor 26 based on the output of each sensor.

  As described above, the walking assist device 10 is worn from the user's thigh to the lower leg. The walking assist device 10 includes a motor 26 that applies torque to the user's leg joints, and sensors 20R, 22R, 24R, 20L, 22L, 24L, and 27 that detect movements of both legs of the user. These sensor groups may be collectively referred to as leg sensors.

  A functional structure of the controller 30 will be described. FIG. 2 shows a block diagram inside the controller 30. FIG. 2 schematically illustrates a part of the function of the controller 30. The controller 30 includes a memory 32, an actual leg trajectory determination module 34, a coefficient determination module 36, a target trajectory determination module 38, a conversion module 40, a difference unit 42, and a motor driver 44. The modules 34, 36, 38, 40, 42, and 44 are the CPU of the controller 30 as hardware, and instructions executed by the CPU are realized as programs.

  First, the reference walking pattern stored in the memory 32 will be described. The reference walking pattern is data describing the trajectory of both legs at the time of walking created under a predetermined condition (reference walking condition). The reference walking condition includes a stride, a walking cycle, and an inclination angle of the ground along the traveling direction. Walking is realized by alternately repeating the same trajectory on both legs. The reference walking pattern describes the trajectory of both legs in which the same curve is repeated alternately on the left and right. The “curve” here may include discontinuous points (points that cannot be differentiated). Specifically, the trajectory of both legs is given as time series data of the relative position of both feet with respect to the hip joint position (or waist position). The trajectory of both legs is represented by two periodic time series data having a predetermined phase difference.

  The reference walking pattern can be divided into a reference free leg trajectory and a reference stance trajectory. The reference free leg trajectory corresponds to the foot trajectory of the free leg period in the reference walking pattern. The reference stance trajectory corresponds to the foot trajectory of the stance period in the reference gait pattern. Specific examples of the reference walking pattern will be described later. In the reference walking pattern, since both legs draw the same trajectory, the reference free leg trajectories of both legs are the same. Similarly, the reference leg trajectories for both legs are the same.

  Next, the actual leg trajectory determination module 34 will be described. A leg sensor signal is input to the actual leg trajectory determination module 34. The actual leg trajectory determination module 34 specifies the trajectories of both legs during walking from the output of the leg sensor. The actual leg trajectory determination module 34 obtains the getting-off timing and the landing timing from the outputs of the ground sensors 24L and 24R. From these timings, the actual leg trajectory determination module 34 specifies the free leg period and the stance period in the actual trajectory. The actual leg trajectory determination module 34 determines the stride WR when the right leg is stepped out and the stride WL when the left leg is stepped out from the obtained actual trajectory of both legs. The actual leg trajectory determination module 34 determines the stride for each step.

  The determined stride lengths WR and WL are input to the coefficient determination module 36. The coefficient determination module 36 stores the past left / right stride ratio WL / WR and calculates a moving average of them. FIG. 2 shows the stride ratio obtained by smoothing AVR (WL / WR). The moving average may be a simple moving average or a weighted moving average. In other words, the coefficient determination module 36 functions as a low-pass filter of the actually measured stride ratio. The moving average of the stride ratio corresponds to the asymmetry coefficient. In FIG. 2, the asymmetry coefficient is represented by the symbol Ce1. Hereinafter, in this embodiment, the asymmetry coefficient is represented by the symbol Ce1. The coefficient determination module 36 outputs the calculated asymmetry coefficient Ce1 to the target trajectory determination module 38.

  The target trajectory determination module 38 reads the reference free leg trajectory of the left leg (affected leg) from the memory 32. As described above, the curve of the reference free leg trajectory is the same for both legs. Here, the “left leg reference swing leg trajectory” was intentionally identified by synchronizing the right leg trajectory in the reference walk pattern with the user's right leg movement, thereby determining the timing of the target leg trajectory of the left leg. It is because it determines.

  Here, the reference free leg trajectory of the left leg is represented by time-series data of the position of the left foot tip relative to the hip joint position. Therefore, the reference free leg trajectory can be represented by a curve having the distance in the front-rear direction of the user as an axis. The target trajectory determination module 38 generates a new trajectory by multiplying the span on the distance axis of the reference free leg trajectory by the asymmetry coefficient Ce1. This new trajectory corresponds to the target free leg trajectory when the left leg is free. In other words, the target trajectory determination module 38 determines a target free leg trajectory obtained by expanding / contracting the reference free leg trajectory along the longitudinal distance axis by the asymmetry coefficient Ce1. The target swing leg trajectory is sent to the conversion module 40.

  The target swing leg trajectory is represented by a time series of positions relative to the hip joint position of the left foot tip. The conversion module 40 converts the time series of the left foot tip position into the time series of the knee joint angles using the joint angles obtained from the leg sensors. A time series of the converted knee joint angles becomes a target knee joint angle, that is, a command value of the motor 26. The conversion module 40 also determines a time-series output timing of the target knee joint angle while collating the actual trajectory identified by the actual leg trajectory determination module 34 with the reference walking pattern, and each time-series value at the determined timing. Is output. That is, the conversion module 40 has a function of synchronizing the left leg target swing leg trajectory with the movement of the right leg.

  The target knee joint angle output by the target trajectory determination module 38 is sent to the subtractor 42. The differentiator 42 calculates the difference between the target knee joint angle and the left knee joint angle detected by the leg sensor. The difference is sent to the motor driver 44. The motor driver 44 controls the motor 26 so that the difference becomes zero.

  As described above, the controller 30 controls the motor 26 so that the left knee joint angle follows the target knee joint angle trajectory. That is, the controller 30 controls the motor 26 so that the movement of the left leg follows the target swing leg trajectory.

  Next, the reference walking pattern and the target swing leg trajectory will be described in detail. FIG. 3 shows a reference walking pattern. The horizontal axis of the graph indicates time, and the vertical axis indicates the position in the advancing direction of the foot tip with the hip joint position as the origin. The upper part of the vertical axis corresponds to the front of the user. The solid line graph indicates the foot leg trajectory of the left leg, and the broken line graph indicates the foot leg trajectory of the right leg. Here, the vertical trajectory of the toes will not be described. Note that the graph of FIG. 3 schematically represents the trajectory, and the graph is actually represented by a curve.

  The state of the left leg is shown above the graph, and the state of the right leg is shown below the graph. Here, the state means either a swinging leg state or a standing leg state. In FIG. 3, the symbol “Sw” means the swinging state, and the symbol “St” means the standing state.

  The meaning of the graph will be described. After leaving the floor at time T1, the right leg swings forward from the hip joint, and then arrives at time T2. That is, the right leg is in the free leg state from time T1 to T2. The left leg is grounded until time T3. From time T2 to T3, the position of the hip joint moves forward with both legs in contact with the ground. That is, the toe positions of both legs move backward relative to the hip joint position while maintaining a certain distance. From time T2 to T3, both legs are in a standing state. The left leg leaves the floor at time T3, then swings forward from the hip joint, and reaches the floor at time T4. That is, the left leg is in the free leg state from time T3 to T4. The right leg is grounded from time T2 to T5. That is, the right leg is in a standing state from time T2 to time T5. The right leg leaves the bed again at time T5. The above operation is repeated.

  In FIG. 3, the shape of the right leg graph and the shape of the left leg graph are the same. That is, in the reference walking data, the trajectories of the foot positions of both legs are the same. When the trajectory of the user's leg is close to the trajectory of the reference walking data, it means that the user has achieved normal walking. Symbol WR indicates the stride of the right leg, and WL indicates the stride of the left leg. In the reference gait data, the right leg stride WR is equal to the left leg stride WL.

  In the left leg graph, the thick line indicates the trajectory of the left leg during the swing leg. That is, the thick line portion of the graph indicates the reference free leg trajectory of the left leg.

  Next, a process for generating a target free leg trajectory from the reference free leg trajectory will be described. As described with reference to FIG. 2, the actual leg trajectory determination module 34 determines the user's right stride WR and left stride WL. The coefficient determination module 36 calculates a moving average of the stride ratio WL / WR. This moving average corresponds to the asymmetry coefficient Ce1. The target trajectory determination module 38 generates a new trajectory obtained by multiplying the span on the distance axis of the reference free leg trajectory by the asymmetry coefficient Ce1. FIG. 4 schematically shows the relationship between the reference free leg trajectory and the generated new trajectory. The meanings of the horizontal and vertical axes of the graph are the same as those in FIG. That is, the vertical axis corresponds to the distance axis in the user front-rear direction. A one-dot chain line indicates the reference trajectory of the left leg. The broken line indicates the reference trajectory of the right leg. In particular, a one-dot chain line in the free leg period corresponds to the reference free leg trajectory. Symbol A1 indicates a range between the maximum value and the minimum value of the reference swing leg period. That is, the reference symbol A1 indicates the distance span of the reference free leg trajectory. The target trajectory determination module 38 reduces the reference free leg trajectory by multiplying the distance span A1 of the reference free leg trajectory by the asymmetry coefficient Ce1. The solid line in FIG. 4 shows a new trajectory reduced by the asymmetry coefficient Ce1. Symbol A2 indicates the distance span of the new trajectory. Here, a relationship of AVR (WL / WR) = Ce1 = A2 / A1 is established. The new trajectory during the free leg period corresponds to the target free leg trajectory.

  The controller 30 controls the motor 26 so as to follow the target swing leg trajectory thus generated. The effect of the target swing leg trajectory will be described. Reference sign B1 in FIG. 4 indicates the stride in the reference free leg trajectory (reference walk pattern). That is, the distance B1 indicates the stride when the healthy leg is stepped one step. Reference B2 indicates the stride in the target swing leg trajectory. The distance B2 in the target swing leg trajectory is smaller than the distance B1. That is, this target free leg trajectory means a trajectory in which the step length B2 of the left leg (affected leg) is smaller than the step length B1 of the right leg (healthy leg). The walking assist device 10 can appropriately assist the movement of the left leg (affected leg) that draws a different trajectory from the right leg (healthy leg) by controlling the actuator so as to follow the target swing leg trajectory. it can.

  The controller 30 changes the asymmetry coefficient Ce1 as the usage time of the walking assist device 10 elapses. FIG. 5 shows changes with time of the asymmetry coefficient Ce1. The graph indicated by the symbol Ce1_s shows the variation of the asymmetry coefficient for each step. Ce1_s is calculated for each step by the actual leg trajectory determination module 34. As shown in the graph, the asymmetry coefficient Ce1_s for each step varies. As described above, the asymmetry coefficient Ce1 corresponds to a moving average of a plurality of past asymmetry coefficients for each step. By calculating the moving average, it is possible to remove the influence of variation at each step. Further, as the walking ability recovers, the user can step forward on the left leg (affected leg). As a result, the asymmetry coefficient Ce1 output from the coefficient determination module 36 also approaches 1 asymptotically. As the asymmetry coefficient Ce1 gradually approaches 1, the difference between the reference free leg trajectory and the target free leg trajectory decreases. As the user's walking ability recovers, the target free leg trajectory automatically approaches the reference free leg trajectory. The walking assist device 10 can adaptively change the assist pattern according to the recovery of the user's walking ability.

  Note that the walking assist device 10 may actively make the asymmetry coefficient Ce1 asymptotic to 1. The walking assist device 10 that executes such processing functions as a rehabilitation device that promotes recovery of the user's walking ability.

  (Second Embodiment) A walking assistance apparatus according to the second embodiment will be described. The walking assist device of the second embodiment is different from the first embodiment in the processing of the controller. The controller of the second embodiment employs the step time ratio as an asymmetry coefficient. FIG. 6 shows a block diagram of the controller 130 of the second embodiment. A portion indicated by a one-dot chain line is a portion different from the controller 30 of the first embodiment. Below, the part shown with the dashed-dotted line is demonstrated.

  The actual leg trajectory determination module 134 determines the actual trajectory of both legs during walking from the output of the leg sensor. The actual leg trajectory determination module 134 obtains the leaving timing and the landing timing from the outputs of the ground sensors 24L and 23R. From those timings, the actual leg trajectory determination module 134 specifies the free leg period and the stance period in the actual trajectory. The actual leg trajectory determination module 34 calculates the required time TR required to step one step on the right leg (healthy leg) and the required time TL required to step one step on the left leg (affected leg) from the obtained actual trajectories of both legs. decide. The actual leg trajectory determination module 134 determines the required time for each step.

  The determined required times TR and TL are input to the coefficient determination module 136. The coefficient determination module 136 stores the past required left / right required time ratio TL / TR and calculates a moving average of them. FIG. 6 shows the stride ratio obtained by smoothing AVR (TL / TR). In this embodiment, the moving average of the required time ratio corresponds to the asymmetry coefficient. In FIG. 6, the asymmetry coefficient is represented by the symbol Ce2. Hereinafter, in this embodiment, the asymmetry coefficient is represented by the symbol Ce2.

  The target trajectory determination module 138 reads the reference free leg trajectory and the reference stance trajectory of the left leg (affected leg) from the memory 32. The reference free leg trajectory has been described in the first embodiment. The reference stance trajectory means the trajectory of the left leg in the range indicated by the stance state St in FIG.

  The target trajectory determination module 138 generates a new trajectory obtained by multiplying the span on the time axis of the reference free leg trajectory by the asymmetry coefficient Ce2. This new trajectory corresponds to the target free leg trajectory when the left leg is free. In other words, the target trajectory determination module 138 determines a target free leg trajectory obtained by expanding and contracting the reference free leg trajectory along the time axis by the asymmetry coefficient Ce2. Similarly, the target trajectory determination module 138 generates a new trajectory obtained by multiplying the span on the time axis of the reference stance trajectory by the reciprocal of the asymmetry coefficient Ce2. This trajectory corresponds to the target stance trajectory. The target swing leg trajectory and the target stance trajectory are sent to the conversion module 40. Subsequent processing is the same as in the first embodiment.

  A process for expanding and contracting the reference free leg trajectory along the time axis will be described. FIG. 7 schematically shows the relationship between the reference free leg trajectory and the generated new trajectory. The meanings of the horizontal and vertical axes of the graph are the same as those in FIG. That is, the horizontal axis corresponds to the time axis. A one-dot chain line indicates the reference trajectory of the left leg. In particular, the one-dot chain line in the period indicated by the symbol TR corresponds to the reference free leg trajectory. The symbol TR indicates the time required for the right leg. That is, the required time TR indicates the time span of the reference free leg trajectory. The target trajectory determination module 138 extends the reference free leg trajectory by multiplying the time span TR of the reference free leg trajectory by the asymmetry coefficient Ce2. The solid line in FIG. 7 shows a new trajectory extended by the asymmetry coefficient Ce2. The symbol TL represents the time span of the new trajectory and corresponds to the required time for the detected left leg. The new trajectory during the free leg period corresponds to the target free leg trajectory.

  This target free leg trajectory means a trajectory in which the required time TL of the left leg (affected leg) is longer than the required time TR of the right leg (healthy leg). The walking assist device 10 appropriately assists the movement of the left leg (affected leg) that cannot be moved quickly like the right leg (healthy leg) by controlling the actuator so as to follow this target swing leg trajectory. can do.

  A process for expanding and contracting the reference stance trajectory along the time axis using the reciprocal of the asymmetry coefficient Ce2 will be described. FIG. 8 shows changes in the state of both legs over time. The horizontal axis in FIG. 8 represents the passage of time. FIG. 8A shows changes with time in the reference walking pattern. FIG. 8B shows changes over time in the target free leg trajectory and the target stance trajectory. FIG. 8A shows the upper and lower views of FIG. 3 again. Reference numerals St and Sw in FIG. 8 represent a standing leg state and a free leg state, respectively. Since FIG. 8A is based on the reference walking pattern, the right required time TR and the left required time TL are equal. In FIG. 8B, the left required time TL is extended by the asymmetry coefficient Ce2. Here, the left required time TL means the time when the left leg is in the swinging leg state, and also means the time TRs when the right leg is in the one leg standing leg state at the same time. On the other hand, reference sign TLs in FIG. 8B means a time during which the left leg is in a single leg standing state. The left leg standing leg time TLs is equal to the right required time (right leg swing leg time TR). As is clear from FIG. 8B, the relationship TLs = TRs × (1 / Ce2) is established. Since the trajectory of the right leg corresponding to time TRs corresponds to the reference leg trajectory, the trajectory obtained by extending the reference leg trajectory by the reciprocal of the asymmetry coefficient Ce2 is ideal for the left leg when the right leg is in the free leg state. This represents the trajectory, that is, the target stance trajectory. The target stance trajectory generated in this way represents an ideal trajectory of the affected leg adapted to the swing leg movement of the healthy leg. The walking assistance device 10 can appropriately assist the movement of the left leg (affected leg) not only in the free leg period but also in the standing leg period by controlling the actuator so as to follow the target stance trajectory.

  In the above description, the period from when the leg leaves the floor to the landing is defined as the free leg period, and the period corresponding to the free leg period in the standing leg is defined as the standing leg period. More suitably, this stance period can be expressed as a “one-leg stance period”. That is, the both-legged leg period exists between the free leg period and the stance period (one legged leg period). Instead of such a definition, it is possible to adopt a definition in which the walking state is divided into a swing leg period and a stance period. Specifically, it is preferable to define the period from the time when the right leg is landed until the next time the left leg leaves the floor as the free leg period of the left leg. This is because both legs are in contact with the ground for a while from the timing when the right leg is landed, but at this time, the right leg supports almost the entire weight and almost no load is applied to the left leg. The left leg swing leg period corresponds to a standing leg period for the right leg. Even if such a definition is adopted, the technique of the above-described embodiment is established.

The points to be noted in understanding the technology of the examples are listed.
(Note 1) The reference walking pattern, that is, the trajectory of the legs, is a relative trajectory with respect to the hips of both feet during walking. The trajectory data may be time series data of the joint angles of both legs. It is well known that there is a unique correspondence between the position of the foot tip relative to the waist and the joint angle of the leg. The walking assist device may store several reference walking patterns. For example, different reference walking patterns may be selected depending on walking conditions and walking environments, such as walking speed, up and down stairs, and up and down slopes. The walking assistance device may generate a reference walking pattern from the movement of the healthy leg (second leg).

(Points to note 2) “Target swing leg with the first leg in the standard walking pattern expanded or contracted by a factor other than 1 (asymmetric factor) along either the time axis or the longitudinal distance axis. “Generating a trajectory” can be translated as “determining a trajectory obtained by multiplying a time span or a distance span of a trajectory in a reference walking pattern by an asymmetry coefficient as a target pattern”.
(Note 3) The actuator may be any actuator that applies torque to at least one of the hip joint, knee joint, and ankle joint of the affected leg. Typically, torque may be applied only to the knee joint of the affected leg. An example of the configuration of such an actuator is as follows. That is, the upper link attached to the thigh of the affected leg and the lower link attached to the lower leg are connected to the actuator so as to be swingable beside the knee joint, and a motor for rotating the lower link is attached to the connecting portion. It has the structure which is made.

  (Note 4) The target free leg trajectory and the target stance trajectory may be expressed as a relative trajectory of the toes with respect to the hip joint. In that case, the controller converts the relative trajectory of the toes into a joint angle trajectory and determines a command value to the actuator. At this time, the actuator is controlled to follow the trajectory of the target joint angle. That is, the controller and the actuator constitute a position control system. The actuator is position controlled, but as a result, torque is applied to the user's joint.

  (Note 5) The reference walking pattern is a trajectory relative to the hips of both feet during walking. The trajectory obtained by deforming the trajectory in the first leg of the reference walking pattern is the target trajectory of the actuator. The controller synchronizes the trajectory of the user's second leg detected by the leg sensor with the trajectory of the second leg in the walking pattern, thereby determining the timing of the target swing leg trajectory with respect to the movement of the user's second leg. For example, the controller synchronizes the target swing leg trajectory with the timing when the user's second leg has landed. In this way, the controller applies torque to the joint of the first leg so that the movement of the first leg follows the target trajectory while synchronizing the target free leg trajectory with the movement of the user's second leg detected by the leg sensor. Add.

  (Note 6) The leg sensor may be an encoder that detects the angle of each joint of the leg. In this case, the leg sensor preferably includes a ground sensor attached to the sole. The movement of the leg can be obtained from the change of the joint angle during the time when the leg is detected by the ground sensor. Alternatively, the leg sensor may be an image sensor that obtains leg movement data from an image of leg movement. Whether or not the leg is a swing leg can also be determined from the image of the leg movement. The leg sensor does not need to detect the movement up to the details of the first leg. For example, leg movement in a direction that intersects the direction of travel is not necessary to practice the present invention.

  (Note 7) In the embodiment, the step ratio and the step time ratio are used as the asymmetry coefficient. By using the stride ratio and the step time ratio at the same time, the reference trajectory in the reference walking pattern may be expanded and contracted along both the time axis and the distance axis. It is also preferable to adopt the ratio of speed when taking a step as the asymmetry coefficient.

  (Note 8) The coefficient determination module functions as a low-pass filter. The cut-off frequency of the low-pass filter is preferably a value smaller than the frequency indicating the recovery degree of walking ability. Since the time required for one step is about 1 second, the cutoff frequency is preferably about 5 Hz at most.

  Specific examples of the present invention have been described in detail above, but these are merely examples and do not limit the scope of the claims. The technology described in the claims includes various modifications and changes of the specific examples illustrated above. The technical elements described in this specification or the drawings exhibit technical usefulness alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing. In addition, the technology exemplified in this specification or the drawings can achieve a plurality of objects at the same time, and has technical usefulness by achieving one of the objects.

The schematic of the walking assistance apparatus of an Example is shown. 1A shows a front view, and FIG. 1B shows a side view. A block diagram of the controller is shown. It is a figure explaining a standard walking pattern. It is a figure explaining a target free leg track. It is a figure which shows the time-dependent change of an asymmetry coefficient. It is a figure which shows the time-dependent change of an asymmetry coefficient. It is a figure explaining the target free leg track | orbit when a step time ratio is employ | adopted. It is a figure which shows the time-dependent change of a walking state.

Explanation of symbols

10: Walking assist device 12R, 12L: Orthosis 14R, 14L: Upper link 16R, 16L: Lower link 20R, 20L: Knee encoder 22R, 22L: Waist encoder 24R, 24L: Ground sensor 26: Motor 27: Tilt sensor 28: Support Bar 30, 130: Controller 32: Memory 34, 134: Actual leg trajectory determination module 36, 136: Coefficient determination module 38, 138: Target trajectory determination module

Claims (3)

The first leg cannot be moved freely, but the second leg is a walking assistance device that assists the walking motion of a healthy user,
A storage device that stores a reference walking pattern that describes the trajectory of the leg during walking in which the trajectories of both legs draw the same curve;
A leg sensor that detects the movement of each leg of the user;
An actuator that is worn by the user and applies torque to the joint of the first leg;
A controller that controls the actuator in synchronization with the movement of the user's second leg,
Based on the output of the leg sensor, to generate a desired free leg trajectory corresponding to the movement of the free leg of the first leg of the user from the reference walking pattern with the stride ratio between the stride of the first leg and the second leg. Determined as the coefficient of
Generating a trajectory at the time of the free leg of the first leg in the reference walking pattern by reducing the trajectory of the first leg along the distance axis in the longitudinal direction of the leg by the coefficient ,
A walking assist device that controls an actuator so that a movement of a user's first leg follows a target swing leg trajectory. The controller
Based on the output of the leg sensor, the time ratio between the time required to step one step on the first leg and the time required to step one step on the second leg is determined as another coefficient.
Walking aid of claim 1, characterized in that stretching in the different coefficients along the time axis the trajectory of the time the free leg of the first leg in the target free leg trajectory. Controller, walking assist device according to claim 1 or 2 with the lapse of use time of the walking assist device is characterized in that is asymptotic to the factor 1.

Images