JP4986658B2 - Motor function support device - Google Patents

Description

  TECHNICAL FIELD The present invention relates to a motor function support device that enhances a motor function for a person whose motor function has deteriorated, for example, a person who has a disability in an upper limb motor function due to spinal cord injury or the like.

  In recent years, an increase in the proportion of elderly people in the population and a decrease in the birth rate have been noticeable, and along with this, it is important to reduce the burden on the caregivers who perform this assistance, as well as helping to reduce motor function due to muscular weakness due to age It has become. In addition, to improve the quality of life (QOL) by encouraging the independence of those who want to receive assistance (persons who are subject to assistance), including those with disabilities in motor functions such as spinal cord injury due to traffic accidents or sports accidents. The movement to plan is also becoming more active.

  For such a situation, a device that assists or assists a motor function by operation of an operator by applying robot technology is generally known.

This conventional apparatus will be described with reference to FIG.
This conventional apparatus is operated by an optional operation of an operator and outputs an operation result information, and a drive for calculating an operation command based on the operation result information and driving a drive mechanism unit 603 described later. A control equipment unit 602 and a drive mechanism unit 603 that is driven by the drive control unit 602 to perform the assistance operation of the person being assisted. The operation device 601 can employ a selection switch in addition to the teaching pendant, and is configured as a master arm when the apparatus is a master / slave robot type.
Here, “optional operation” means an operation that causes a mechanical action directly on the external environment due to the movement of, for example, a limb based on the operator's own intention ( And so on)

  In such a conventional apparatus, the operation result information output from the operation device 601 is processed by the drive control unit 602 to generate an operation command. The time required for this process is usually on the order of several ms. Thus, an operation command can be generated substantially in real time, and the drive mechanism unit 603 can operate. Therefore, as long as the operator operates the operation device 601, the operator can operate the drive unit 603 to reflect the operator's intention.

  When a selection switch is used as the operation device 601, an operation pattern stored in advance in the drive control unit 602 is output as the operation command based on operation result information that is selected by the selection switch. Driving the drive mechanism unit 603 is also performed. In this case, after the operation pattern is selected by the operator's selection switch, the drive mechanism unit 603 automatically operates.

  However, in such a conventional apparatus, when a teaching pendant or a master arm is used as the operation device 601, the driving unit 603 does not operate unless the operator arbitrarily operates the operation device 601. That is, only when the operator's intention is confirmed, the operation device 601 is voluntarily operated and the mechanism driving unit 603 operates. Therefore, a delay always occurs with respect to the operator's intention and the assistance operation is executed. Will be. Further, when the operator's intention is changed, a greater delay may occur. Further, in the case of a person having a disorder in the upper limb motor function due to spinal cord injury or the like, the operation device 601 itself is operated. Often difficult to do.

  Further, when a selection switch is employed as the operation device 601, an assisting operation corresponding to a predetermined operator's intention is possible, but an immediate response is made when the operator's intention changes during the assisting operation. I was disliked that I couldn't work, but rather took time to respond.

  In addition, there is an apparatus that estimates an operator's intention to exercise from an electromyogram or an electrooculogram and performs a prosthetic hand movement for a person with a disability in an upper limb motor function (for example, Patent Document 1). However, Patent Document 1 does not preempt the operator's intention prior to the operator's optional operation, as in the prior art.

  Here, there is a result that suggests that the possibility of estimation of the intention is correlated with the selection judgment by a psychological experiment and that the prediction that precedes it can be estimated from the eye movement. The knowledge that there is a possibility that a preceding latent process may be detected from biological information is known (for example, Non-Patent Document 1).

  As a device related to this knowledge, devices (driving behavior intention estimation system) such as Patent Literature 2 and Patent Literature 3 that estimate an operator's intention from eye movement are known.

This apparatus will be described more specifically with reference to FIG.
This apparatus includes means 701 for measuring / analyzing the line-of-sight trend of the operator (driver), means 702 for measuring a situation other than the operator, and driving action intention estimating means 703 for estimating the intention of the operator. It is prepared for. Further, the means 701 for measuring / analyzing the visual line trend of the operator includes a visual line direction detecting means 704 and a visual line direction frequency distribution calculating means 705. Further, the means 702 for measuring a situation other than the operator includes an environment information measuring means 706, a vehicle behavior data measuring means 707, and a driving operation amount measuring means 708.

  In this apparatus, after the gaze direction detection means 704 detects the gaze direction of the operator, the gaze direction frequency distribution calculation means 705 calculates the gaze trend frequency distribution and sends the result to the driving action intention estimation means 703. On the other hand, the environment information measuring unit 706 measures the road surface and the like, the vehicle behavior data measuring unit 707 measures the moving state of the vehicle, and the driving operation amount measuring unit 708 measures an arbitrary operation amount such as a steering wheel. The result is transmitted to the driving action intention estimation means 703, respectively. Based on these results, the driving action intention estimation means 703 estimates the operator's intention or predicts the future driving situation and supports the operation by limiting driving or the like. That is, in this apparatus, the operator's operation intention can be estimated from the operator's line-of-sight trend and optional operation (in this case, driving of the vehicle).

JP-A-5-111881 Japanese Patent Publication No. 2002-331850 JP 2005-174218 A Shinsuke Shimojo, “Where does the activeness of actions go ?: Neural mechanisms of decision making”, Journal of the Japan Society of Mechanical Engineers, Vo.109, No.1049, pp.261-264, 2006

  However, in the driving behavior intention estimation systems such as Patent Documents 3 and 4, the operator's voluntary operation is estimated because the operation intention is estimated based on the result of the voluntary operation of the operator (in this case, driving the vehicle). It is difficult to estimate the future operating state when it is necessary to input the operation result, when the operation is stopped without any optional operation, or when the optional operation is difficult due to the motor function disorder for the operator. Therefore, there is a problem that it is difficult to perform an operation that anticipates the operator's intention prior to the operator's voluntary operation.

  Furthermore, the prediction calculation processing in the psychological experiment of Non-Patent Document 1 is post-hoc (offline), and it is only described that there is a technical problem about real-time (online) processing. There is no mention of technical matters that link the estimation result to the operation of the motor function support device.

  Therefore, in view of the above-described circumstances, the present invention estimates the intention under the operator's consciousness in order to realize an operation that responds quickly even when the operator's intention changes in the middle of the operation. It is an object of the present invention to provide a motor function support device that can perform an operation that anticipates the operator's intention prior to the above.

The object of the present invention is achieved by the following configurations (1) to (9).
(1) Drive mechanism means for supporting the motor function of the operator, operation means that is directly operated based on the operator's intention and outputs the operation result as operation result information, and a command from the operator A drive control means for generating drive commands based on the drive mechanism means to perform drive control based on the motor function support device ,
Furthermore,
A selection confirmation means that is operated based on the intention of the operator and outputs a selection confirmation result for confirming the operation command;
A biological motion measuring means for measuring the biological motion of the operator and outputting the measurement result as biological motion information;
A biological motion analysis means for generating a motion analysis result based on the selection confirmation result and the biological motion information;
And includes, and said drive control means, based on said operation result information and the selection decision result and the motion analysis result, the operation command to generate, motor function assisting device for driving and controlling the drive mechanism unit In
The biological movement measuring means measures at least eye movement as the biological movement, and when measuring the eye movement, it measures the sight line angles in the vertical and horizontal directions in time series, and the biological movement analyzing means comprises: A motor function support device that determines that the line of sight is stopped when a time during which the time-series change in the line-of-sight angle is within a predetermined range is equal to or greater than a preset threshold value.
(2) The biological motion analysis means includes:
(I) The biological motion information accumulated in a predetermined period is read in a batch, and based on the read biological motion information, the misclassification probability is obtained when the threshold of the gaze stop time is used as a judgment boundary. Set the threshold of discrimination score from the target value of another probability,
(II) After calculating the discrimination score from the measured biological motion information,
(III) Determine gaze retention based on the threshold value and the discrimination score based on the biological motion information, and, as a result of this discrimination, generate the motion analysis result based on the gaze angle at which gaze retention is determined, And (IV) classifying the biological motion information sequentially read from the biological motion measuring means into two types of data according to the length of time of the eye-gaze retention, and accumulating it for a predetermined period of time,
(V) Based on the accumulated data, the threshold value and the misclassification probability are obtained,
The motor function support apparatus according to (1) , wherein the above (II) to (V) are sequentially performed.
(3) The biological motion analysis means includes:
When classifying into two types of data according to the length of time of the gaze stop,
When there is no operation of the operation means, the time series data of the line-of-sight angle from the time when it is newly read after the previous classification until the time when it is determined as the next line-of-sight stop point,
When there is an operation of the operation means, the data is classified into two types of data: time-series data of line-of-sight angles from the time when the operation means is started until the time when the operation means is started after the previous classification. The motor function support device according to (2) above, characterized in that:
(4) The biological motion analysis means includes:
When adjusting the threshold so that the misclassification probability takes a preset target value,
The motor function support apparatus according to (2) or (3) above, wherein a discrimination score corresponding to a predetermined target value of a false discrimination probability is obtained and the discrimination score is set as a threshold value.
(5) The biological motion analysis means includes:
When adjusting any one of the operation command speed, the operation command acceleration, and the cutoff frequency corresponding to the target value of the misclassification probability,
In order to obtain the same effect as increasing the target value of the misclassification probability, the operation command speed or the operation command acceleration is decreased according to a preset ratio, or the cut-off frequency is decreased and adjusted,
In order to obtain an effect equivalent to lowering the target value of the misclassification probability, the adjustment is performed by increasing the operation command speed or the operation command acceleration or increasing the cut-off frequency according to a preset ratio. The motor function support device according to any one of (2) to (4) , wherein
(6) The biological motion analysis means includes:
When generating the motion analysis result,
Based on the line of sight and the position of the work object set in advance,
A virtual straight line in a three-dimensional space representing the line of sight is calculated, a virtual sphere having a predetermined radius centered on the center of gravity of the work target is set, and the work target is set at the intersection of the virtual straight line and the virtual sphere. The motion analysis result is generated so as to select a work object including a point closest to the center of gravity position of the object and closest to the operator's eyeball. (2) to (5) above The motor function support apparatus according to any one of the above.
(7) A switch is used as the operation means,
The drive control means includes
Determining whether or not the operation means and the selection confirmation means are operated,
When determining that there is no operation of the operation means and operation of the selection confirmation means, generate an operation command based on the motion analysis result,
When it is determined that there is an operation of the selection confirmation means, the operation waits until the switch is operated. The motor function support device according to any one of (1) to (6), wherein an average is calculated and an operation command is generated based on the calculation result.
(8) A joystick with a bar is used as the operation means,
The drive control means includes
Determining whether or not the operation means and the selection confirmation means are operated,
When determining that there is no operation of the operation means and operation of the selection confirmation means, generate an operation command based on the motion analysis result,
When it is determined that there is an operation of the selection confirmation means, the operation waits until the joystick is operated, and when the joystick bar is operated, an operation command is issued based on the direction and angle information to tilt the field bar. The motor function support device according to any one of the above (1) to (6) , wherein
(9) The biological movement measuring means measures at least a head movement as the biological movement,
The motor function support apparatus according to any one of (1) to (8) , wherein the head posture angle is measured when the head movement is measured.

  According to the configuration of (1) above, an operation command for the drive mechanism unit is generated based on the biological motion information of the operator and an optional operation, so that the operator can quickly respond to the operator's intention and operate in the middle of the operation. Even when the operator's intention changes, an action that responds promptly is realized, so that the operator's intention can be anticipated prior to the operator's voluntary operation.

According to the above configuration (1), further, by making a determination of the operator's gaze angle measured gaze stopping, it is possible to obtain a biological motion information more reflecting an intention of the operator.

According to the configuration of (2) above, it is possible to perform discrimination corresponding to changes in biometric information over time by obtaining the threshold of eye-gaze stop time and the misclassification probability using the biomotion information accumulated for a certain period. As a result, it is possible to anticipate the intention under the operator's consciousness.

According to the configuration of (3) above, the threshold value and the misclassification probability can be sequentially updated by sequentially classifying the data into groups based on whether or not the operation means is operated.

According to the configuration of (4) above, the misclassification probability can be made substantially constant by adjusting the threshold value of the eye-gaze stop time in accordance with the misclassification probability. Can be preempted.

According to the configuration of (5) above, by adjusting the operation command speed, the operation command acceleration, or the cut-off frequency according to the misclassification probability, an intentional action such as a rapid operation of the drive mechanism means even when misclassification is performed. It is possible to prevent inconveniences contrary to the above.

According to the configuration of (6) above, a virtual straight line in the three-dimensional space representing the line of sight is calculated, and a virtual sphere having a predetermined radius with the center of gravity of the work object as a center is set. By generating the motion analysis result so as to select the work object including the point closest to the center of gravity position of the work object and the closest distance from the eyeball of the operator among the intersections with the virtual sphere, the discrimination result Accordingly, it is possible to generate an operation pattern of the drive mechanism unit for reliably selecting a work object corresponding to the work object and approaching the work object, thereby improving the accuracy of the operation of the drive mechanism unit.

According to the configuration of the above (7) and (8) , the operation command of the drive mechanism unit is output according to the motion analysis result or the operation result information depending on the presence / absence of operation of the operation means, so that the operator's intention can be flexibly Corresponding operations can be executed immediately and operability can be improved. Further, since a switch or a joystick is used as the operation means, the operation can be simplified.

  According to the present invention, in order to realize an immediate action even when the operator's intention changes in the middle of the operation, the intention under the operator's consciousness is estimated, and the operation is performed prior to the operator's voluntary operation. It is possible to perform an operation in anticipation of a person's intention.

  Embodiments according to the present invention will be described below with reference to the drawings.

Hereinafter, one embodiment of the motor function support device according to the present invention will be described. FIG. 1 is a block diagram of the motor function support device according to the present invention, FIG. 2 is a schematic diagram of the motor mechanism support device according to the present invention, and FIG. 3 is an operation flowchart of the biological motion analysis unit according to the present invention. FIG. 4 is an operation flowchart of the drive control unit according to the present invention.
In addition, the embodiment described below will be described as an example applied to a meal operation of the operator for a person whose upper limb motor function is restricted due to spinal cord injury (hereinafter also referred to as an operator), The present invention is not limited to this, and includes operations that can support existing motor functions (for example, shopping operations, tool gripping operations, etc.), including healthy persons, in addition to elderly people with decreased motor functions. The present invention can be applied to various operations without departing from the above.

  As shown in FIG. 1, an exercise mechanism support apparatus 100 according to the present embodiment includes a biological motion measurement unit (biological motion measurement unit) 101, a biological motion analysis unit (biological motion analysis unit) 102, and an operation device (operation unit). 103, a drive control unit (drive control unit) 104, a drive mechanism unit (drive mechanism unit) 105, and a selection confirmation switch (selection confirmation unit) 106. As shown in FIG. 2, the operator 1 sits in a wheelchair in front of a table on which food 2 (hereinafter also referred to as a work object) is placed, and on the back of the wheelchair. It is assumed that the drive mechanism unit 105 is attached so as not to interfere with the operator 1.

The biological movement measuring unit 101 measures at least one biological movement of the eye movement of the operator 1, head movement, movement of the head or shoulder arm, myoelectric potential of the head or shoulder arm, and the like. The measurement result is output as biological motion information.
In the present embodiment, it is assumed that the meal operation of the operator 1, that is, the food selection operation is performed. Therefore, the eye movement of the operator 1 is adopted as the biological information. That is, the biological movement measuring unit 101 according to the present embodiment is configured by an eye movement measuring device such as an eye mark recorder (manufactured by Nack Image Technology), and in this case, when the operator looks at food. In addition, the visual angle information in the vertical and horizontal directions is measured in time series and output as the biological motion information.

  In addition, the biological motion analysis unit 102 receives the biological motion information (in this embodiment, line-of-sight angle information), performs analysis based on the biological motion information, and uses the analysis result as the motion analysis result. Output to. In addition, the biological motion analysis unit 102 determines that the line of sight is stopped when the time during which the time series change in the line-of-sight angle information is within a predetermined range is equal to or greater than a preset threshold value K. When the motion analysis result is generated, a selection confirmation result, which is an output signal of a selection confirmation switch 106 described later, is input and used as feedback information.

The operation device 103 is a terminal that the operator 1 optionally operates. For example, an angle meter composed of a potentiometer such as a joystick, or an arm or head motion measuring instrument using a magnetic sensor or an optical sensor. The information such as the angle by these sensors is output to the drive control unit 104 as operation result information. The operation device 103 also outputs the operation result information as feedback information to the biological motion analysis unit 102.
Note that a switch may be employed as the operation device 103 to simplify the operation. In this case, for example, the drive control unit 104 determines that the line of sight is stopped before and after the moment the switch is pressed. The motor function support device 100 may be configured to calculate an average of the line-of-sight angle at the time of being generated and to generate an operation command.

The drive control unit 104 generates an operation command for the drive mechanism unit 105 based on the motion analysis result and the operation result information. At this time, the operator 1 selects or confirms the operation command by the selection confirmation switch 103, and the drive control unit 104 receives the result and selects based on either the motion analysis result or the operation result information. Therefore, an operation command is output (details will be described later).
In accordance with this operation command, the drive mechanism unit 105 is controlled and driven.

The drive mechanism unit 105 is configured as an articulated manipulator that is partly attached to the forearm of the operator's dominant arm in order to support the operator's 1 meal operation. In response, an assisting force is applied to the operator 1 to perform an arm assistance operation of the operator 1.
The drive mechanism unit 105 is not limited to this, and various mechanisms can be employed. For example, a scalar type manipulator mechanism, a PUMA type manipulator mechanism, etc. that can perform all human eating operations instead of the operator 1. In this case, a meal utensil such as a spoon or fork is attached as an effector. At this time, the drive mechanism unit 105 approaches and grips the food 2 and carries the food 2 to the mouth of the operator 1 to assist the operator's meal operation.

  Next, the operation flow of the biological motion measuring unit 101 will be described in more detail with reference to FIG.

  First, the biological motion analysis unit 101 collectively reads biological motion information accumulated in a predetermined period (in this embodiment, line-of-sight angle information) (that is, S401), and based on the read biological motion information, A general formula of the misclassification probability when the threshold of the stop time is used as a discrimination boundary is obtained, and the threshold is set from the target value of the misclassification probability (ie, S402). Note that the misclassification probability is generally a discriminant analysis which is a kind of statistical processing, and when applied to the present invention, the misclassification probability is a probability of erroneously judging the length of the stopping time with respect to the threshold value K.

ここでｚは判別スコア、ａ＝［ａ１．．．ａｎ］はｎ次元判別係数ベクトル、ａ０は定数、ｘ＝［ｘ１．．．ｘｎ］は前記生体運動情報のｎ次元データベクトルである。判別係数ベクトルは、各群に属する複数の計測値から得られる平均値ベクトルと不偏共分散行列とから、推定値として求められる。このとき、群Ａを選択中のデータ、群Ｂを確定のデータとする。 Here, the derivation of the misclassification probability will be specifically described.
It is assumed that the population of the measured values follows an n-dimensional normal distribution, and all groups to which the measured values belong have the same population covariance matrix. Here, n-dimensions indicate a generalized n-dimensional data vector that corresponds to a plurality of measurement objects such as a line-of-sight angle or a head movement stop time, for example.

Here, z is a discrimination score, and a = [a1. . . an] is an n-dimensional discrimination coefficient vector, a0 is a constant, and x = [x1. . . xn] is an n-dimensional data vector of the biological motion information. The discrimination coefficient vector is obtained as an estimated value from an average value vector obtained from a plurality of measurement values belonging to each group and an unbiased covariance matrix. At this time, the group A is selected data and the group B is determined data.

となる。ここで、Σは不偏共分散行列で|Σ|はその行列式である。また、Ｄは平均値ベクトルと不偏共分散行列Σとから求められるマハラノビスの汎距離である。なお、

と標準化している。 Then, when determining which group the data x belongs to with respect to the determination score z obtained for a specific measurement data x, the data belonging to one group is erroneously determined as the other group. The general formula of the probability, that is, the misclassification probability p is

It becomes. Here, Σ is an unbiased covariance matrix and | Σ | is its determinant. Further, D is the Mahalanobis generalized distance obtained from the mean value vector and the unbiased covariance matrix Σ. In addition,

It is standardized.

Further, the description will be continued by returning to FIG. 3 again.
The living body motion measurement unit 101 obtains a corresponding discrimination score from a predetermined target value of the misclassification probability by the equations (2) and (3), and sets the discrimination score as a threshold value K. That is, first, from the above equation (2), the target value of the misclassification probability is substituted as the initial setting value, and the discrimination score satisfying equation (2) is set as the threshold value K. At this time, if the misclassification probability to be substituted is lowered, the threshold value of the discrimination score is lowered (that is, the threshold value of the gaze stop time is raised), and the probability of generating an operation command for the drive mechanism unit 105 is lowered. As a result, the drive mechanism unit 105 operates more reliably, and the drive mechanism unit 105 can be prevented from moving accidentally against the intention of the operator 1.
Note that the operation pattern of the drive mechanism unit 105 may be adjusted without adjusting the threshold value K. In other words, setting the upper limit of speed, acceleration, and cut-off frequency has the same effect as raising the target value of misclassification probability, and conversely, setting the lower limit is equivalent to lowering the target value of misclassification probability. effective.

  Next, a discrimination score is calculated from the measurement data, that is, biological motion information (that is, S404). Substituting the measurement data x into the above equation (1), the discrimination score z is obtained.

  Further, the biological motion analysis unit 102 performs the determination by comparing the threshold value K with the determination score z, and as a result of the determination, the movement analysis result is obtained based on the line-of-sight angle when it is determined that the line-of-sight is stationary. Generate (ie, S404). That is, the motion analysis result is generated based on the line-of-sight angle when the line-of-sight is stopped and the position of the work object 2 set in advance.

In this generation, for example, a virtual straight line in the three-dimensional space representing the line of sight of the operator 1 is calculated based on the line-of-sight angle, line-of-sight stop time, and work object position, and the center of gravity of the work object is centered. A virtual sphere having a predetermined radius is set, and a work object including a point closest to the center of gravity of the work object and the closest distance from the eyeball of the operator among the intersections of these virtual straight lines and the virtual sphere A trapezoidal velocity pattern for selecting and approaching the work object is generated as the motion analysis result. However, the discrimination score z and the threshold value K are compared, and when the discrimination score z is smaller than the threshold value K, the operation pattern is output, and conversely, when it is larger, the operation pattern is not used and the drive mechanism is used. The unit 105 is put on standby.
In addition, as a method for identifying the coordinate system of the eyeball of the operator 1 and the coordinate system of the work object, various known methods such as a method using an optical position measuring device or a method using a camera can be used. .

  The biological motion analysis unit 102 generates the motion analysis result, and then reads the line-of-sight angle information from the biological motion measurement unit 101 and the selection confirmation result from the selection confirmation switch 103, and the length of time for the line-of-sight stoppage. According to the above, the data is classified into two types of data of Group A and Group B and accumulated for a predetermined period. Based on the accumulated data, the threshold value K for the eye stop time and the misclassification probability are obtained, and the process returns to S403. Repeat the operation. That is, the biological motion analysis unit sequentially performs this series of operations (ie, S403 to S406) at a predetermined sampling period (for example, several ms).

Further, regarding the two types of data classification methods of the A group and the B group, the group A is set as the set of the line-of-sight stop time being selected, the group B is set as the set of the line-of-sight stop time at the time of determination, and the selection determination switch 106 The data is classified into groups according to whether or not the operation is performed.
More specifically, as the group A, the operation device 103 is not operated and the selection confirmation switch 106 is not operated. After the previous classification, the group A moves to the next line-of-sight stop point. The time series data of the line-of-sight stop time until the point of time is accumulated. Further, as the group B, time-series data of the line-of-sight stop time from the time of the new classification to the time when the operation of the selection confirmation switch 106 is started after the previous classification is accumulated. At this time, in order to obtain the misclassification probability in the operation of S403, the average vector and the unbiased co-division matrix are obtained from the data vectors belonging to the group A and the group B.

Further, the operation flow of the drive control unit 104 will be described with reference to FIG.
The drive control unit 104 obtains operation result information from the operation device 103 (ie, S501), a maneuver analysis result from the biological motion analysis unit 102 (ie, S502), and a selection confirmation result from the selection confirmation switch 106 ( That is, S503), respectively. Next, it is determined whether or not the operation device 103 and the selection confirmation switch 106 are operated (that is, S504). When it is determined that neither the operation device 103 is operated nor the selection confirmation switch 106 is operated, a motion analysis is performed. An operation command is generated based on the result (ie, S506).
However, when it is determined that the operation device 103 is operated, the motion command is generated in a form in which the motion analysis result and the operation result information are superimposed. That is, the operation result information is added to the position information based on the motion analysis result and output as a final motion command (ie, S507).

  On the other hand, if it is determined that there is an operation of the selection confirmation switch, the operation waits until the operation of the operation device 103, and generates an operation command based on the operation result information when the operation device is operated (that is, S508).

  After the operation command is generated, the operation command is output to the drive mechanism unit 105 (that is, S508), and the operations of S501 to S508 are sequentially repeated at a predetermined sampling period (for example, several ms).

  The drive control unit 104 generates such an operation command, so that the operator 1 can operate the operation device 103 while using the operation command based on the motion analysis result as a reference for the movement of the mechanism drive unit 105 to the food 2. Fine adjustment can be made in substantially real time by the operation of. When the mechanism driving unit 105 moves in the wrong direction, the operator 1 operates the selection confirmation switch 106 to set the operation mode of the operation device 103 only, and operates the operation device 103 to correct the trajectory. Can do.

  As described above, according to the present embodiment, it is possible to promptly respond to the operator's intention by generating the operation command of the drive mechanism unit based on the biological motion information of the operator and the optional operation. In addition, even when the operator's intention changes during the operation, a quick operation can be realized, so that the operator's intention can be anticipated prior to the operator's optional operation. Furthermore, by measuring the operator's line-of-sight angle and determining the line-of-sight stop, it is possible to obtain biological motion information that more reflects the operator's conscious intention.

  Further, according to the present embodiment, it is possible to perform the determination corresponding to the change in the biological information with time by obtaining the threshold value K of the stationary time and the erroneous determination probability using the line-of-sight information accumulated for a certain period.

Then, according to the biological information analysis unit of the present embodiment, the determination corresponding to the temporal change of the biological information is performed by obtaining the threshold K of the stationary time and the erroneous determination probability using the line-of-sight information accumulated for a certain period. Can do. Further, by sequentially classifying data into groups based on whether or not the operation device is operated, the threshold value K and the misclassification probability can be updated sequentially. In addition, the misclassification probability can be made substantially constant by adjusting the stop time threshold value K according to the misclassification probability, and the speed / acceleration cut-off frequency of the motion pattern in the motion analysis result can be misclassified. By adjusting according to the above, even when erroneous determination is made, the influence of the malfunction of the drive mechanism section can be minimized, and the safety of the operator can be ensured.
Furthermore, according to the biological motion analysis unit of the present embodiment, the intersection between the virtual straight line in the three-dimensional space representing the line of sight and the virtual sphere representing the vicinity of the work target is closest to the center of gravity of the work target and Since the work object including the point closest to the eyeball is selected, the work object corresponding to the determination result can be reliably selected.

  Furthermore, according to the drive control unit of the present embodiment, the operation command of the drive mechanism unit is output according to the motion analysis result or the operation result information depending on whether or not the operation device and the selection confirmation switch are operated. It is possible to execute an action that corresponds quickly in substantially real time.

  This is the end of the description of the specific embodiment. However, the aspect of the present invention is not limited to this embodiment, and appropriate modifications, improvements, and the like are possible. For example, in the present embodiment, most of the processing is performed by software, but a part or all of the processing may be realized by hardware such as FPGA (Field Programmable Gate Array).

  Further, in the aspect of the present invention, the case where eye movement is adopted as the biological movement has been described as an example, but the present invention is not limited thereto, and head movement may be measured. For example, as shown in FIG. 5, it may be configured so that a biological motion measuring unit configured by a head motion sensor (201) that can output angle information such as a gyro sensor is mounted on the operator's head, Even in this case, the above embodiment can be applied and the same effects can be obtained. Furthermore, if the eye movement and the head movement are used together as the biological movement, the accuracy of the operation command of the drive mechanism (105) can be improved.

It is a block block diagram of the motor function support apparatus which concerns on this invention. It is the schematic of the exercise | movement mechanism assistance apparatus which concerns on this invention. It is an operation | movement flowchart of the biological motion analysis part which concerns on this invention. It is an operation | movement flowchart of the drive control part which concerns on this invention. It is the schematic of the exercise | movement mechanism assistance apparatus of another modification based on this invention. It is a schematic block diagram of the conventional apparatus. It is a schematic block diagram of another conventional apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Motor function support apparatus 101 Biological motion measurement part (biological motion measurement means)
102 biological motion analysis unit (biological motion analysis means)
103 Operation device (operation means)
104 Drive control unit (drive control means)
105 Drive mechanism section (drive mechanism means)
106 Selection confirmation switch (selection confirmation means)

Claims (9)

Drive mechanism means for supporting the motor function of the operator, operation means that is directly operated based on the operator's intention, and outputs the operation result as operation result information, and based on a command from the operator and drive control means for controlling the driving and generating an operation command of the drive mechanism unit, a motor function support apparatus having a
Furthermore,
A selection confirmation means that is operated based on the intention of the operator and outputs a selection confirmation result for confirming the operation command;
A biological motion measuring means for measuring the biological motion of the operator and outputting the measurement result as biological motion information;
A biological motion analysis means for generating a motion analysis result based on the selection confirmation result and the biological motion information;
And includes, and said drive control means, based on said operation result information and the selection decision result and the motion analysis result, the operation command to generate, motor function assisting device for driving and controlling the drive mechanism unit In
The biological movement measuring means measures at least eye movements as the biological movements, and measures the eye angles in the vertical and horizontal directions in time series when measuring the eye movements, and
The movement function support device, wherein the biological movement analysis means determines that the line of sight is stopped when the time series change in the line of sight is within a predetermined range is equal to or greater than a predetermined threshold. . The biological motion analysis means includes
(I) The biological motion information accumulated in a predetermined period is read in a batch, and based on the read biological motion information, the misclassification probability is obtained when the threshold of the gaze stop time is used as a judgment boundary. Set the threshold of discrimination score from the target value of another probability,
(II) After calculating the discrimination score from the measured biological motion information,
(III) Determine gaze retention based on the threshold value and the discrimination score based on the biological motion information, and, as a result of this discrimination, generate the motion analysis result based on the gaze angle at which gaze retention is determined, and
(IV) The biological motion information sequentially read from the biological motion measuring means is classified into two types of data according to the length of time of the gaze stop and accumulated for a predetermined period,
(V) Based on the accumulated data, the threshold value and the misclassification probability are obtained,
The motor function support apparatus according to claim 1, wherein (II) to (V) are sequentially performed . The biological motion analysis means includes
When classifying into two types of data according to the length of time of the gaze stop,
When there is no operation of the operation means, time series data of the line-of-sight angle from the time of the new classification to the time of determining that the next line-of-sight is stopped after the previous classification,
When there is an operation of the operation means, the data is classified into two types of data: time-series data of line-of-sight angles from the time when the operation means is started until the time when the operation means is started after the previous classification. motor function support device according to claim 2, characterized in that. The biological motion analysis means includes
When adjusting the threshold so that the misclassification probability takes a preset target value,
The motor function support device according to claim 2 or 3 , wherein a discrimination score corresponding to a predetermined target value of a false discrimination probability is obtained, and the discrimination score is set as a threshold value . The biological motion analysis means includes
When adjusting any one of the operation command speed, the operation command acceleration, and the cutoff frequency corresponding to the target value of the misclassification probability,
In order to obtain the same effect as increasing the target value of the misclassification probability, the operation command speed or the operation command acceleration is decreased according to a preset ratio, or the cut-off frequency is decreased and adjusted,
In order to obtain an effect equivalent to lowering the target value of the misclassification probability, the adjustment is performed by increasing the operation command speed or the operation command acceleration or increasing the cut-off frequency according to a preset ratio. The motor function support apparatus according to any one of claims 2 to 4, wherein The biological motion analysis means includes
When generating the motion analysis result,
Based on the line of sight and the position of the work object set in advance,
A virtual straight line in a three-dimensional space representing the line of sight is calculated, a virtual sphere having a predetermined radius centered on the center of gravity of the work target is set, and the work target is set at the intersection of the virtual straight line and the virtual sphere. any of claims 2-5, characterized in that to generate the motion analysis results to select the work object including the points in the closest from the nearest and the operator's eye to the center-of-gravity position of the object 1 The motor function support device according to item . As the operation means, a switch is used,
The drive control means includes
Determining whether or not the operation means and the selection confirmation means are operated,
When determining that there is no operation of the operation means and operation of the selection confirmation means, generate an operation command based on the motion analysis result,
When it is determined that there is an operation of the selection confirmation means, the operation waits until the switch is operated, and when the switch is operated, the line-of-sight angle when it is determined that the line-of-sight is stopped before and after the moment. by calculating the average, motor function support device according to any one of claims 1 to 6, characterized in that for generating an operation command based on the calculation result. As the operation means, using a joystick with a bar ,
The drive control means includes
Determining whether or not the operation means and the selection confirmation means are operated,
When determining that there is no operation of the operation means and operation of the selection confirmation means, generate an operation command based on the motion analysis result,
When determined that there is an operation of the selection decision unit, it waits until there is an operation of the joystick, when the bars of the joystick is operated, operation command based on the direction and angle information to defeat the field bar motor function assistance device according to claim 1, characterized in that to produce a. The biological movement measuring means measures at least head movement as the biological movement,
At the time of measurement of head movement, the motor function assistance device according to any one of claims 1-8, characterized in that for measuring the attitude angle of the head.

Images