JP5849473B2 - Robot, robot control method, and program - Google Patents

Description

  The present invention relates to a robot, a robot control method, and a program thereof.

  In an environment where people and robots coexist and live and work, observation of the operation status of the robot (especially detection of abnormal movement) is an important factor for safety management. For example, there is a method of monitoring a torque value of a motor that drives a robot and issuing a warning when a specified value is exceeded. Further, in Patent Document 1, the difference between the amount of work performed by the robot drive unit on the robot and the estimated energy change held by the robot is detected, and an alarm is issued when the difference exceeds a preset threshold. We have proposed a method and a control device that emits and stops the robot.

JP 2010-188504 A

  However, in the conventional methods as described above, identification of whether or not the system is operating normally and countermeasures when an abnormality is detected (alarm or drive stop) are performed, but there is a problem that no further information can be obtained. It was. For example, the method and the control device described in Patent Document 1 have a problem that only single determination information of normality or abnormality with respect to a value calculated from work amount or energy can be captured. Specifically, there are various stages of robot operating conditions. Even if an abnormal state occurs, abnormal load (abnormality that is larger than expected), collision / impact (impact due to abnormal operation or collision), equipment backlash, etc. However, there is a problem in that it is impossible to identify the abnormality due to various factors such as a setting error. As a result, there is a problem that information such as the cause of abnormality and the content of abnormality cannot be obtained, and it takes time to take appropriate measures.

  SUMMARY An advantage of some aspects of the invention is to solve at least a part of the problems described above, and the invention can be implemented as the following forms or application examples.

  [Application Example 1] A robot according to this application example detects an inertial sensor attached to an operation part of the robot, a drive part that drives the operation part, a control part that controls the drive part, and an operation state of the robot. And a detection unit that transmits a detection result to the control unit, wherein the detection unit detects an operation state of the robot from a signal pattern based on an output signal from the inertial sensor, and the control unit is a drive unit based on the detection result It is characterized by controlling.

  According to this application example, since the operation state of the robot is detected from the signal pattern based on the output signal from the inertial sensor attached to the operation unit of the robot, it is possible to identify abnormalities and state differences due to various factors. .

  Application Example 2 In the robot according to the application example described above, the detection unit detects the operation state of the robot by comparing a comparison reference prepared in advance with a signal pattern.

  According to this application example, since the operation status of the robot is detected by comparing the comparison reference provided in advance with the signal pattern, various identifications of the operation status can be easily performed. Specifically, comparison criteria for various operating situations assumed in advance are created, and various identifications of operating conditions can be made by verifying the similarity to the signal pattern.

  Application Example 3 In the robot according to the application example described above, the comparison criterion includes a plurality of abnormality determination criteria, and the detection unit detects any abnormality determination criterion in light of the plurality of abnormality determination criteria. Is detected, the abnormality detection result corresponding to the abnormality determination criterion is transmitted to the control unit, and the control unit controls the drive unit based on the abnormality detection result.

  According to this application example, when it is detected that the signal pattern corresponds to one of the abnormality determination criteria in light of the plurality of abnormality determination criteria, the corresponding abnormality detection result is transmitted to the control unit. Therefore, it is possible to easily identify various abnormalities in the operation status and perform control according to the abnormalities. Specifically, various comparisons (abnormality determination standards) for various operational abnormalities assumed in advance are created, and various similarities of abnormal operational situations can be identified by verifying the similarity to the signal pattern. That is, when an abnormality occurs, information such as the cause of the abnormality and the content of the abnormality is obtained, and a more appropriate response is possible.

  Application Example 4 In the robot according to the application example described above, the comparison criterion includes a plurality of load determination criteria, and the detection unit detects any one of the load determination criteria in light of the plurality of load determination criteria. Is detected, the load detection result corresponding to the load determination criterion is transmitted to the control unit, and the control unit controls the drive unit based on the load detection result.

  According to this application example, when it is detected that the signal pattern corresponds to any one of the load determination criteria in light of the plurality of load determination criteria, the corresponding load detection result is transmitted to the control unit. Therefore, it is possible to easily perform load identification and control according to the load. Specifically, it is possible to detect a difference in load in normal operation by creating a comparison standard (load judgment standard) for an assumed operating load in advance and verifying the similarity with the signal pattern. it can. For example, since the load difference due to the weight (inertial mass) and size (moment of inertia) of the transfer object handled by the robot can be detected, the contents of the handling work can be changed depending on the transfer object. It becomes.

  [Application Example 5] In the robot according to the application example described above, the plurality of load determination criteria include one or both of a determination criterion for each load amount and a determination criterion for each target load. Prepared in advance as a set of corresponding signal patterns obtained for each standard load for a plurality of known standard loads, and the judgment criteria for each target load is for each target load for a plurality of types of known target loads. It is characterized in that it is prepared in advance as a set of corresponding signal patterns obtained in the above.

  According to this application example, the comparison criterion includes a plurality of load determination criteria, and the plurality of load determination criteria includes a corresponding signal pattern obtained for each standard load with respect to a plurality of standard loads with known load amounts. A set or a set of corresponding signal patterns obtained for each target load is prepared in advance for a plurality of types of known target loads. Therefore, the load can be easily identified. Specifically, load judgment criteria (determination criteria by load amount) are prepared for each weight of the transfer object handled by the robot (for example, every 100 g) or for each moment of inertia. It is possible to identify an object to be transferred. In addition, for example, when the type of transfer object to be handled is known in advance and the weight and size of each object are different, by preparing a load determination standard (determination standard for each target load) for each type Therefore, it is possible to identify the type of transfer object being handled during operation.

  Application Example 6 In the robot according to the application example described above, the signal pattern is a frequency characteristic pattern of vibration that captures vibration of the operating unit.

  According to this application example, the signal pattern is a frequency characteristic pattern of vibration that captures the vibration of the moving part, so the difference in the state of the moving part of the robot and the inertial mass and moment of inertia of the transfer target to be handled. Differences can be easily identified. Specifically, a difference in state due to various factors such as a load difference generated in the operating part, an impact, a backlash of the apparatus, a setting error, and the like appears as a difference in vibration amplitude and frequency in the operating part. By detecting this frequency characteristic pattern with an inertial sensor attached to the operating part, the state of the operating part can be detected and identified more easily.

  Application Example 7 A robot control method according to this application example includes a step of taking an output signal from an inertial sensor attached to an operating part of the robot, a step of generating a signal pattern based on the output signal, and a robot from the signal pattern. And a step of generating a detection result by detecting the operation state of the device and a step of controlling a drive unit that drives the operation unit based on the detection result.

  According to this application example, since the operation state of the robot is detected from the signal pattern based on the output signal from the inertial sensor attached to the operation unit of the robot, it is possible to identify abnormalities and state differences due to various factors. . Specifically, since the inertial sensor can detect the degree of acceleration, impact, vibration, rotation, etc., it depends on various factors such as the difference in load generated in the operating part, impact, device backlash, setting error, etc. Differences in state can be detected. As a result, when an abnormality occurs, expected information such as the cause of the abnormality and the content of the abnormality can be obtained, a more appropriate response is possible, and a difference in load during normal operation can also be detected.

  Application Example 8 In the robot control method according to the application example described above, a detection result is generated from a detection result obtained by comparing a comparison reference prepared in advance with a signal pattern.

  According to this application example, since the operation status of the robot is detected by comparing the comparison reference provided in advance with the signal pattern, various identifications of the operation status can be easily performed. Specifically, comparison criteria for various operating situations assumed in advance are created, and various identifications of operating conditions can be made by verifying the similarity to the signal pattern.

  Application Example 9 In the robot control method according to the application example described above, the comparison criterion includes a plurality of abnormality determination criteria, and the signal pattern corresponds to one of the abnormality determination criteria in light of the plurality of abnormality determination criteria. Then, when detected, a detection result is generated as a corresponding abnormality detection result, and the drive unit is controlled based on the abnormality detection result.

  According to this application example, when a signal pattern is detected as corresponding to one of the abnormality determination criteria in light of a plurality of abnormality determination criteria, a detection result is generated as a corresponding abnormality detection result, and the abnormality detection is performed. Based on the result, the drive unit is subjected to abnormality handling control. Therefore, it is possible to easily identify various abnormalities in the operation status and perform control according to the abnormalities. More specifically, various comparisons for abnormal operation situations can be identified by preparing comparison criteria for various operation abnormalities assumed in advance and verifying the similarity to the signal pattern. That is, when an abnormality occurs, information such as the cause of the abnormality and the content of the abnormality is obtained, and a more appropriate response is possible.

  Application Example 10 In the robot control method according to the application example described above, the comparison criterion includes a plurality of load determination criteria, and the signal pattern corresponds to one of the load determination criteria in light of the plurality of load determination criteria. Then, when detected, a detection result is generated as a corresponding load detection result, and the drive unit is controlled based on the load detection result.

  According to this application example, when it is detected that the signal pattern corresponds to one of the load determination criteria in light of a plurality of load determination criteria, a detection result is generated as a corresponding load detection result, and load detection is performed. Based on the result, the drive unit is subjected to load control. Therefore, it is possible to easily perform load identification and control according to the load. Specifically, it is possible to detect a difference in load in a normal operation by creating a comparison standard for a presumed operating load and verifying the similarity with the signal pattern. For example, since the load difference due to the weight (inertial mass) and size (moment of inertia) of the transfer object handled by the robot can be detected, the contents of the handling work can be changed depending on the transfer object. It becomes.

  [Application Example 11] In the robot control method according to the application example described above, the plurality of load determination criteria include both or any one of a load amount determination criterion and a target load determination criterion. Prepared in advance as a set of corresponding signal patterns obtained for each standard load with respect to a plurality of standard loads with known load amounts, and the judgment criteria for each target load is a target for a plurality of types of known target loads. A set of corresponding signal patterns obtained for each load is prepared in advance.

  According to this application example, the plurality of load determination criteria include a determination criterion for each load amount provided in advance based on a plurality of loads, or a determination criterion for each target load provided in advance based on a plurality of predetermined target loads. Since it is either, the load can be easily identified. Specifically, load judgment criteria (determination criteria by load amount) are prepared for each weight of the transfer object handled by the robot (for example, every 100 g) or for each moment of inertia. It is possible to identify an object to be transferred. In addition, for example, when the type of transfer object to be handled is known in advance and the weight and size of each object are different, by preparing a load determination standard (determination standard for each target load) for each type Therefore, it is possible to identify the type of transfer object being handled during operation.

  Application Example 12 In the robot control method according to the application example described above, the signal pattern is a frequency characteristic pattern of vibration that captures vibration of the operating unit.

  According to this application example, since the signal pattern is a frequency characteristic pattern in which the moving part vibrates, the difference in the state of the moving part of the robot and the difference in the inertial mass and moment of inertia of the transfer object to be handled can be easily achieved. Can be identified. Specifically, a difference in state due to various factors such as a load difference generated in the operating part, an impact, a backlash of the apparatus, a setting error, and the like appears as a difference in vibration amplitude and frequency in the operating part. By detecting this frequency characteristic pattern with an inertial sensor attached to the operating part, the state of the operating part can be detected and identified more easily.

  Application Example 13 A program according to this application example is characterized by causing a robot to function including the control method described above.

  According to this application example, by using the program including the control method described above to function, the robot's operation due to various factors such as the difference in load generated in the robot's operating part, impact, device backlash, setting error, etc. The operating state can be detected. Therefore, when an abnormality occurs, expected information such as the cause of the abnormality and the content of the abnormality can be obtained, and a difference in load during normal operation can be detected. As a result, it is possible to provide a robot that can perform an appropriate treatment and response according to the operating state and abnormality more easily and quickly.

FIG. 3 is a perspective view illustrating a configuration of the robot according to the first embodiment. FIG. 2 is a block diagram illustrating a functional configuration of the robot according to the first embodiment. Explanatory drawing which shows the basic principle of an operating condition detection. The graph which shows the example of the frequency characteristic pattern in stop operation | movement. The graph which shows the example of the frequency characteristic pattern by an impact. The graph which shows the example of the angular velocity attenuation | damping pattern in stop operation | movement. 3 is a flowchart for explaining the operation of the robot according to the first embodiment.

  DESCRIPTION OF EMBODIMENTS Embodiments embodying the present invention will be described below with reference to the drawings. The following is one embodiment of the present invention and does not limit the present invention. In the following drawings, the scale may be different from the actual scale for easy understanding.

(Embodiment 1)
FIG. 1 is a perspective view illustrating a configuration of a robot 100 according to the first embodiment.
As shown in FIG. 1, the robot 100 is a transfer device for components and the like, and includes a machine base 10, an arm shaft portion 20, an arm drive motor 21, a feeding / discharging material arm 30, a hand drive mechanism 40, a holding hand 50, and integration. It is comprised from the control part 60, the acceleration sensor 70, the angular velocity sensor 71, etc.

The machine base 10 incorporates an arm drive motor 21 and a bearing mechanism (not shown).
One end (lower part) of the arm shaft portion 20 is supported by the machine base 10 and freely rotates around the rotation shaft by the arm drive motor 21 and the bearing mechanism.
One end of the feeding / discharging material arm 30 is fixed to the end (upper part) of the arm shaft portion 20 opposite to the side supported by the machine base 10.
The supply / discharge material arm 30 rotates around the rotation axis of the arm shaft portion 20.
A hand drive mechanism 40 is fixed to the end of the supply / discharge material arm 30 opposite to the side fixed to the arm shaft portion 20.
A holding hand 50 as an operating unit is attached to the hand drive mechanism 40.
The holding hand 50 has a function of grabbing and releasing a transfer object such as a component, and can be moved up and down and rotated by the hand drive mechanism 40.

  That is, the robot 100 uses the arm drive motor 21 and the hand drive mechanism 40 as a drive unit to bring the holding hand 50 into and out of contact with the transfer object, and the transfer object held by the holding hand 50 is placed on the mounting place. Lifting up or placing it on the mounting place.

Further, the holding hand 50 is provided with an acceleration sensor 70 and an angular velocity sensor 71 as inertia sensors.
The acceleration sensor 70 can detect the acceleration due to the rotation of the supply / discharge material arm 30 or an external impact.
The angular velocity sensor 71 can detect an angular velocity when the supply / discharge material arm 30 is rotated and an angular velocity when the holding hand 50 is rotated.

The integrated control unit 60 performs overall control of the operation of each unit of the robot 100 based on a control program input in advance via an information input / output device (not shown) such as a personal computer. The integrated control unit 60 includes a detection unit 61 and a control unit 62.
The detection unit 61 receives output signals from the acceleration sensor 70 and the angular velocity sensor 71 and notifies the control unit 62 of the detection result.
Upon receiving the detection result, the control unit 62 issues a command to the arm drive motor 21, the hand drive mechanism 40, and the holding hand 50 to control the operation.

Next, the functional configuration of the robot 100 will be described with reference to FIG.
FIG. 2 is a block diagram illustrating a functional configuration of the robot 100 according to the first embodiment.
As described above, the robot 100 includes the holding hand 50 as an operating unit, and the arm driving motor 21 and the hand driving mechanism 40 as a driving unit that moves the holding hand 50. Further, the holding hand 50 includes an acceleration sensor 70 and an angular velocity sensor 71 as inertial sensors.
The holding hand 50 also has a function as a drive unit that performs operations such as grasping and releasing the transfer object.

The control unit 62 issues an operation command to the arm drive motor 21, the hand drive mechanism 40, and the holding hand 50 based on the control program.
The arm drive motor 21, the hand drive mechanism 40, and the holding hand 50 receive a command and perform a desired operation.
The acceleration sensor 70 and the angular velocity sensor 71 transmit the detection results of acceleration and angular velocity generated by the movement and operation of the operating unit (holding hand 50) to the detection unit 61 as output signals.
The detection unit 61 analyzes the output signal, and detects the operation status of the operation unit by comparing the obtained signal pattern with a comparison standard provided in advance as a database.
The detection unit 61 transmits the detection result of the operation status to the control unit 62.
The control unit 62 that has received the detection result issues a command corresponding to the detection result to the arm drive motor 21, the hand drive mechanism 40, and the holding hand 50 based on the control program.

Next, a method for detecting the operating status of the operating unit will be described.
FIG. 3 is an explanatory diagram showing the basic principle of operation status detection.
The robot 100 has a structure having a hand drive mechanism 40 and a holding hand 50 at the tip of the supply / discharge material arm 30 with the arm shaft portion 20 as a fulcrum. Therefore, when the spring constant of the supply / discharge material arm 30 is k and the total load inertia of the supply / discharge material arm 30, the holding hand 50, and the hand drive mechanism 40 is j, as shown in the one-mass model of FIG. The robot 100 has a natural vibration with a frequency ω given by ω ² = k / j. This natural vibration can be observed when the mass point j is freely vibrated. That is, when the operating part (holding hand 50) having the arm shaft part 20 as a fulcrum freely vibrates, this vibration can be observed by the inertial sensor (acceleration sensor 70) attached to the operating part.

  On the other hand, in the actual operating state, the operating part vibrates variously while receiving various influences based on the natural vibration. For example, vibration while turning after the feed / release material arm 30 starts turning from a stopped state, vibration immediately after the feed / release material arm 30 stops, before and after the hand drive mechanism 40 moves the holding hand 50, Alternatively, vibration during movement, vibration after the holding hand 50 lifts or places the transfer object, and the like. The frequency of these vibrations varies depending on the position of the holding hand 50 and the weight of the object to be held. In addition, the vibration amplitude, decay time, and the like vary depending on the operation status (operation of the robot 100). However, in a similar state, condition, and situation, patterns such as vibration frequency and amplitude are observed as certain natural vibration patterns. Therefore, it is possible to detect the operating status of the operating unit by analyzing the signal sensed by the acceleration sensor 70 attached to the operating unit.

  In addition, the operating status can be captured not only by the above-described detection by vibration but also by a change in angular velocity in the rotation and rotation of the holding hand 50. Therefore, it is possible to detect the operating status of the operating unit by analyzing a signal sensed by the angular velocity sensor 71 attached to the operating unit. In implementation, it is preferable to use the acceleration sensor 70 and the angular velocity sensor 71 in a complementary manner.

Specifically, the detection of the operation status by these inertial sensors is performed by, for example, pre-recording the output signal of the inertial sensor in a series of normal operations as a reference signal pattern, and from the inertial sensor observed in actual operation. The actual operation status can be evaluated by comparing the output signal with the reference signal pattern each time. The evaluation of the operation status is an evaluation of the degree of difference between the desired operation and the actual operation status. In addition, by preparing a reference signal pattern for an assumed abnormal operation in advance, it is possible to evaluate the similarity with an assumed abnormal operation situation. An assumed abnormality is a load abnormality experienced by an operating part.For example, when an object that is heavier than expected is held, an impact caused by an abnormal operation or collision, etc., an abnormality due to equipment play or a setting error, etc. For example, when there is an action.
With such an evaluation, it is possible to identify normality / abnormality of the operation status, analogy of the abnormal state / error cause at the time of abnormality, analogy and identification of the magnitude of the load during normal operation, and the control of the robot 100 associated therewith. Is also possible.

FIG. 4 is a graph showing an example of the frequency characteristic of acceleration detected by the acceleration sensor 70 in the stopping operation of the supply / discharge material arm 30.
In the graph of FIG. 4, the solid line indicates the frequency characteristic of the acceleration in the stop operation when the holding hand 50 has nothing, and the broken line indicates the holding operation when the holding hand 50 has a 2 kg object.
As can be seen from the portion indicated by the one-dot chain line circle in the graph, the characteristics of the vibration frequency at the time of stopping vary greatly depending on the load.

FIG. 5 is a graph showing an example of an acceleration frequency characteristic pattern detected by the acceleration sensor 70 when an impact is applied.
In the graph of FIG. 5, the solid line indicates the frequency characteristic of acceleration during normal operation, and the broken line indicates the acceleration frequency characteristic when a light impact is applied.
As can be seen from the portion indicated by the one-dot chain line in the graph, two peaks are observed at 10 to 100 Hz, and the vibration frequency at the time of impact is detected.

FIG. 6 is a graph showing an angular velocity attenuation pattern detected by the angular velocity sensor 71 in the stopping operation of the supply / discharge material arm 30.
In the graph of FIG. 6, the solid line indicates the attenuation pattern of the angular velocity in the stop operation when the holding hand 50 has nothing, and the broken line indicates the holding operation when the holding hand 50 has a 2 kg object.
As can be seen from the portion indicated by the one-dot chain line in the graph, the damping characteristics at the time of stopping vary greatly depending on the load.

  As described above, in the actual operation state, the operation unit is subjected to various vibrations while being affected in various ways, and the difference in various states can be detected by the inertial sensor attached to the operation unit. . Specifically, it is possible to determine whether the operation status is normal or abnormal, to detect the content of the abnormality in the case of an abnormality, and to identify the size and weight of the load in the case of a normal operating range. . Hereinafter, a specific method for evaluating the actual operation status such as detection and identification will be described.

1. Preparation 1.1. Setting the timing and type of operating status detection As a preparation stage, first, the timing for evaluating and detecting the operating status and the type of evaluation are set. The timing of evaluation is, for example, a predetermined period from the time when a command is issued when the supply / discharge material arm 30 is largely rotated, a predetermined period after the stop operation command, a predetermined period during which the holding hand 50 is rotating, etc. It is. The types of evaluation include identification / evaluation of normal / abnormal operating conditions, analogy evaluation of abnormal conditions / causes of abnormalities, analogy evaluation of the magnitude of load during normal operation, and the like.
These setting results are reflected in the control program and controlled by the integrated control unit 60.

1.2. Creation of Comparison Criteria Next, comparison criteria corresponding to the timing and type of each evaluation are created. The comparison criteria include operation criteria for identifying normal / abnormal operating conditions, abnormality criteria for estimating abnormal conditions and causes of abnormalities, and load criteria for estimating the magnitude of loads during normal operation. and so on. These comparison criteria are stored as a database and referred to as needed for comparison when the robot 100 is operated (FIG. 2).

1.2.1. Operation Judgment Criteria The operation judgment criteria include signal patterns (normal operation patterns) indicating respective normal operations and threshold information for determining normality / abnormality in accordance with various evaluation timings.

  The reference operation pattern by the normal operation is created by operating the robot 100 in an actually normal state and acquiring the output signal of the inertia sensor at the set evaluation timing. Specifically, for example, the output signal from the acceleration sensor 70 is subjected to FFT transform (Fast Fourier Transform fast Fourier transform) to create a frequency characteristic pattern (amplitude pattern for each frequency).

The threshold information for determining normality / abnormality is set as a threshold value of the similarity between the reference operation pattern of the operation determination criterion and a signal pattern observed in actual operation (similarly a pattern subjected to FFT conversion). Specifically, the determination is performed by setting a fixed threshold value to the normalized cross-correlation value indicating the similarity of each pattern.
Normalized cross correlation (NCC) is an absolute measure normalized by an average and a standard deviation. A value closer to 1 indicates a higher pattern similarity. Further, by using the normalized cross-correlation value, the similarity of the pattern shape is evaluated without being affected by the difference in amplitude.

Note that the evaluation of similarity is not limited to normalized cross-correlation when evaluating differences in amplitude or the like, but may be simply evaluation based on cross-correlation.
Further, as described above, a method may be used in which evaluation is continuously performed over a series of work operations without subdividing the evaluation timing. That is, this is a method in which a signal pattern which is a reference over a series of work operations prepared in advance and an output signal of an inertial sensor during actual operation are continuously compared. In this case, when the degree of similarity of the pattern exceeds a certain range, it is compared and evaluated with the abnormality determination criterion, and the abnormality content is detected.

1.2.2. Abnormality Determination Criteria The abnormality determination criterion includes a signal pattern indicating an abnormality assumed at various timings (assumed abnormality pattern), and threshold information for determining whether or not applicable.

The assumed abnormal pattern is a method created by operating the robot 100 in an actually assumed abnormal state and acquiring the output signal of the inertial sensor, or a signal obtained in an experiment different from the operation of the robot 100. Create it by a method such as using it. Further, the assumed abnormal pattern is created as a frequency characteristic pattern by FFT conversion, for example, similarly to the operation determination standard.
In addition, threshold information for determining whether or not it is applicable is also set as a normalized cross-correlation value or a threshold for the cross-correlation value, similarly to the operation determination criterion.
1.2.3. Load Judgment Criteria The load judgment criteria are composed of signal patterns (reference load patterns) assumed for various loads at various timings and threshold information for performing appropriate / non-applicable determination.

The reference load pattern is created by obtaining the output signal of the inertial sensor by operating the robot 100 in a state where several known loads are applied, that is, in a state where a transfer object having a known weight and shape is handled. It is created by a method or a method using a signal obtained in an experiment different from the operation of the robot 100. The reference load pattern is created, for example, as a frequency characteristic pattern by FFT conversion as in the case of the operation determination reference.
In addition, threshold information for determining whether or not it is applicable is also set as a normalized cross-correlation value or a threshold for the cross-correlation value, similarly to the operation determination criterion.

  In addition, as a way of holding the assumed reference load pattern, for example, a method of preparing as a determination criterion for each load amount with respect to loads of a plurality of reference weights and reference moments of inertia, or a transfer object to be handled by the robot 100, for example, There is a method of preparing a judgment criterion for each target load for each type of part or product.

1.3. Preparation of command Next, a command operation as a result of detecting the operation status is set in advance. Specifically, when an abnormality is detected, a warning is issued or the robot 100 is stopped, or when a specific load is detected, a command corresponding to the load is issued. Is set in advance. The setting content of the command is controlled by the integrated control unit 60 by reflecting it in the control program.

2. Detection of Operation Status FIG. 7 is a flowchart for explaining the operation of the robot 100.
Next, the flow of detecting the operation status of the robot 100 will be described with reference to the flowchart of FIG. 7 and FIG.
First, the operating conditions of the robot 100 are set (step S1). Specifically, the mechanical setup of the robot 100 and the setting of the control program are performed. The setting of the control program includes setting of comparison criteria (selection of comparison criteria prepared in advance, etc.).

Next, the robot 100 is started (step S2), and the first operation is performed (step S3). Further, the output signals of the acceleration sensor 70 and the angular velocity sensor 71 generated in accordance with the operation are taken into the detection unit 61 (step S4).
Next, it is confirmed whether the operation determination is necessary at this timing, that is, whether it is set as the operation determination timing (step S5). If it is not set as the operation determination timing, the next step (step S7) is performed. Proceed to If operation determination is necessary, operation determination is performed (step S6). Specifically, the detection unit 61 generates a signal pattern by, for example, performing FFT conversion on the captured output signals of the acceleration sensor 70 and the angular velocity sensor 71, and compares it with a comparison reference (operation determination reference) specified in this operation. And evaluate whether it is normal or abnormal.

As a result of the operation determination, when it is determined that there is an abnormality, an abnormality determination is performed (step E1). Specifically, in the detection unit 61, the signal pattern is compared with designated comparison criteria (a plurality of assumed abnormality judgment criteria), which abnormality judgment criteria are met, or which abnormality judgment criteria. It is determined whether it is not applicable.
Next, the detection unit 61 transmits an abnormality detection result to the control unit 62 according to the result of the abnormality determination, and the control unit 62 issues an abnormality response command according to the content of the abnormality detection result to the arm drive motor 21 and the hand drive mechanism. 40, to the holding hand 50, etc. (step E2). Here, you may perform the treatment which issues a warning etc. as needed.

  If it is determined as normal as a result of the operation determination (step S6), it is confirmed whether the load determination is necessary at this timing, that is, whether the load determination timing is set (step S7). If not set, the process proceeds to the next step (step S10). If load determination is necessary, load determination is performed (step S8). Specifically, the detection unit 61 generates a signal pattern by, for example, performing FFT conversion on the captured output signals of the acceleration sensor 70 and the angular velocity sensor 71 (if it has already been generated in the operation determination, it needs to be generated again) In comparison with the comparison criteria (a plurality of load judgment criteria) specified in this operation, it is judged which of the load judgment criteria is met or which of the load judgment criteria is not met. More specifically, for example, when determining the level of the weight of the transfer object to be handled by the holding hand 50, the determination is performed by comparing with a determination criterion for each load amount with respect to loads of a plurality of reference weights. Further, for example, when determining the parts and product types of the transfer object to be handled by the holding hand 50, the determination is performed by comparing with the judgment criteria for each target load with respect to loads for each of the plurality of parts and product types.

  Next, the detection unit 61 transmits the load detection result to the control unit 62 according to the load determination result, and the control unit 62 issues a load correspondence command according to the content of the load detection result to the arm drive motor 21 and the hand drive mechanism. 40, to the holding hand 50 or the like (step S9). Specifically, for example, a work command is issued such as sorting transfer objects by weight or sorting by part or product type.

  Next, the integrated control unit 60 confirms whether the operation of the robot 100 is completed according to the control program (step S10), and ends the operation if completed. When the subsequent operation is performed, the process proceeds to the next operation (step S11), and a series of designated operations are repeatedly performed (step S3).

As described above, according to the robot, the robot control method, and the program according to the present embodiment, the following effects can be obtained.
Since the operation state of the robot 100 can be detected by evaluating the signal pattern based on the output signals from the inertial sensors (acceleration sensor 70, angular velocity sensor 71) attached to the holding hand 50 of the robot 100, abnormalities and states due to various factors can be detected. The difference can be identified. Specifically, since the acceleration sensor 70 and the angular velocity sensor 71 can detect the degree of acceleration, impact, vibration, rotation, etc., the difference in load generated in the holding hand 50, impact, device backlash, setting error, etc. A state due to various factors such as can be detected. As a result, when an abnormality occurs, expected information such as the cause of the abnormality and the content of the abnormality can be obtained, a more appropriate response is possible, and a difference in load during normal operation can also be detected.

Further, since the operation status of the robot 100 is detected by comparing a comparison reference prepared in advance with a signal pattern, various identifications of the operation status can be easily performed. Specifically, comparison criteria for various operating situations assumed in advance are created, and various identifications of operating conditions can be made by verifying the similarity to the signal pattern.
For example, it is possible to identify various abnormal operation situations by preparing a comparison reference reflecting various abnormal states assumed in advance and verifying the similarity with the signal pattern. That is, when an abnormality occurs, information such as the cause of the abnormality and the content of the abnormality is obtained, and a more appropriate response is possible.
In addition, for example, a comparison reference for a presumed operating load is created, and a difference in load in normal operation can be detected by verifying the similarity with the signal pattern. For example, since a load difference due to a difference in weight (inertial mass) or size (moment of inertia) of a transfer object handled by the robot 100 can be detected, the type of transfer object can be identified, It can be used to change the contents of handling work.

  Further, when the signal pattern is captured as a frequency characteristic pattern in which the holding hand 50 vibrates, it is possible to easily identify the difference in the state of the operating part of the robot 100 and the difference in the inertial mass and the inertial moment of the transferred object to be handled. Can do. Specifically, a difference in state due to various factors such as a load difference generated in the operating part, an impact, a backlash of the apparatus, a setting error, and the like appears as a difference in vibration amplitude and frequency in the operating part. By detecting this frequency characteristic pattern with an inertial sensor attached to the operating part, the state of the operating part can be detected and identified more easily.

  As described above, according to the robot, the robot control method, and the program according to the present embodiment, not only the normality / abnormality of the robot, but also the load identification, abnormal load (abnormality that is larger than expected), collision・ Because it can identify various operating conditions such as shocks (impacts caused by abnormal operations or collisions), equipment play, setting errors, etc., the robot can perform appropriate work corresponding to the load and appropriate measures without taking time to deal with abnormalities. Can be provided.

  DESCRIPTION OF SYMBOLS 10 ... Machine stand, 20 ... Arm shaft part, 21 ... Arm drive motor, 30 ... Feeding / discharging material arm, 40 ... Hand drive mechanism, 50 ... Holding hand, 60 ... Integrated control part, 61 ... Detection part, 62 ... Control part , 70 ... acceleration sensor, 71 ... angular velocity sensor, 100 ... robot.

Claims (9)

An inertial sensor attached to the moving part of the robot,
A drive unit for driving the operating unit;
A control unit for controlling the driving unit;
A detection unit that detects an operation status of the robot and transmits a detection result to the control unit;
The detection unit corresponds to the load determination criterion when a signal pattern based on an output signal from the inertial sensor is detected as corresponding to any one of the load determination criteria in light of a plurality of load determination criteria. The load detection result is transmitted to the control unit,
The said control part controls the said drive part based on the said load detection result, The robot characterized by the above-mentioned. Before Symbol detection unit, wherein the signal pattern, in light of the failure determination reference of several, if it is detected that either of the failure determination reference, the abnormality detection result corresponding to the failure determination reference control unit Communicate to
The robot according to claim 1 , wherein the control unit controls the driving unit based on the abnormality detection result. The plurality of load determination criteria are composed of both or one of the determination criteria according to load amount and the determination criteria according to target load,
The determination criterion for each load amount is prepared in advance as a set of corresponding signal patterns obtained for each standard load for a plurality of standard loads with known load amounts,
The target load another criterion, the plural types of known object load was to claim 1 or 2, characterized in previously prepared is that as a set of the corresponding signal pattern obtained for each of the target load was The robot described. The robot according to any one of claims 1 to 3 , wherein the signal pattern is a frequency characteristic pattern of vibration that captures vibration of the operating unit. Capturing an output signal from an inertial sensor attached to the moving part of the robot;
Generating a signal pattern based on the output signal;
Transmitting the load detection result corresponding to the load determination criterion to the control unit when it is detected that the signal pattern corresponds to any one of the load determination criteria in light of a plurality of load determination criteria ;
And a step of controlling a drive unit that drives the operating unit based on the detection result. Before SL signal pattern, in light of the failure determination reference of several, if it is detected that either of the failure determination reference, the detection result, generates a corresponding abnormality detection result,
The robot control method according to claim 5 , wherein the drive unit is controlled based on the abnormality detection result. The plurality of load determination criteria are composed of both or one of the determination criteria according to load amount and the determination criteria according to target load,
The determination criterion for each load amount is prepared in advance as a set of corresponding signal patterns obtained for each standard load for a plurality of standard loads with known load amounts,
The target load another criterion, the plural types of known object load was to claim 5 or 6, characterized in previously prepared is that as a set of the corresponding signal pattern obtained for each of the target load was The robot control method described. The signal pattern is a control method for a robot according to any one of claims 5 to 7, characterized in that a frequency characteristic pattern of vibrations that capture the vibration of the moving part. A program for causing a robot to function including the control method according to any one of claims 5 to 8 .

Images