JP2015078884A - Monitoring device, monitoring program, and monitoring method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect abnormality of a moving part without being limited to predetermined movements.SOLUTION: A monitoring device comprises: an estimation unit which, on the basis of movement contents of a moving part included in a target device and characteristics how the target device transmits vibration generated by movements of the moving part, estimates characteristics of vibration generated in the target device; and a detection unit which detects abnormality of the moving part by comparing the estimated vibration characteristics and measured vibration characteristics of the target device.

Description

  The present invention relates to a monitoring device, a monitoring program, and a monitoring method for monitoring the operation of a movable part incorporated in the device.

  Machine tools, robots, and the like have movable parts such as servo motors that generate driving force and bearings that transmit the generated driving force. For example, a sensor for detecting acceleration, vibration, etc. is arranged around each movable part such as a servo motor, and each movable part is obtained by comparing the vibration data measured by each sensor with the normal data acquired in advance. Have been proposed (see, for example, Patent Documents 1 and 2).

JP-A-5-281020 International Publication No. 2010/010688

  However, in the prior art, the vibration data at the normal time of the movable part is acquired by each sensor by causing the movable part to perform a predetermined operation. Therefore, the movable part operates differently from the predetermined operation. In this case, it may be difficult to detect abnormality of the movable part.

  In one aspect, an object of the present invention is to detect an abnormality in a movable part without being limited to a predetermined operation.

  A monitoring device according to one aspect estimates a characteristic of vibration generated in a target device based on a characteristic that the target device transmits vibration generated by an operation of a movable part included in the target device and an operation content of the movable part. And a detection unit that detects an abnormality of the movable unit by comparing the estimated vibration characteristic and the measured vibration characteristic of the target device.

  A monitoring program according to another aspect estimates a characteristic of vibration generated in a target device based on characteristics of the target device transmitting vibration generated by movement of the movable part included in the target device and operation contents of the movable part. Then, by comparing the estimated vibration characteristic with the measured vibration characteristic of the target device, the computer is caused to execute a process of detecting an abnormality of the movable part.

  According to another aspect of the monitoring method, the characteristics of the vibration generated in the target device are estimated based on the characteristics of the target device transmitting the vibration generated by the operation of the movable part included in the target device and the operation content of the movable part. The detection unit detects an abnormality of the movable part by comparing the estimated vibration characteristic with the measured vibration characteristic of the target device.

  The abnormality of the movable part can be detected without being limited to the predetermined operation.

It is a figure which shows one Embodiment of a monitoring apparatus. It is a figure which shows the example of the monitoring process in the monitoring apparatus shown in FIG. It is a figure which shows another embodiment of the monitoring apparatus. It is a figure which shows the example of the characteristic of the operation function of the movable part shown in FIG. 3, operation | movement of a movable part, and the measured vibration of the target apparatus. It is a figure which shows the example of the monitoring process in the monitoring apparatus shown in FIG. It is a figure which shows another embodiment of the monitoring apparatus. It is a figure which shows the example of the frequency distribution of the operation function when the three movable parts shown in FIG. 6 operate | move, and the object apparatus. It is a figure which shows the example of the time change of the weighting coefficient calculated | required by the detection part shown in FIG. It is a figure which shows the example of the monitoring process in the monitoring apparatus shown in FIG. It is a figure which shows another embodiment of the monitoring apparatus. It is a figure which shows the example of the monitoring process in the monitoring apparatus shown in FIG. It is a figure which shows the example of the hardware constitutions of the monitoring apparatus shown in FIG.1, FIG.3, FIG.6 and FIG.

  Hereinafter, embodiments will be described with reference to the drawings.

  FIG. 1 shows an embodiment of a monitoring device.

  A monitoring device 100 illustrated in FIG. 1 includes an estimation unit 10 and a detection unit 20. The monitoring device 100 is connected to the monitoring target device MA and the output device 4 via, for example, an interface unit included in the monitoring device 100 in a wired or wireless manner. The target device MA is a machine tool, a robot, or the like, and includes a control unit 1, a movable unit 2, and a measurement unit 3.

  The control unit 1 functions, for example, when a processor or the like included in the target device MA executes a program or the like, and outputs a control signal for controlling the operation to the movable unit 2. Further, the control unit 1 receives a response to the output control signal from the movable unit 2. The control unit 1 outputs the control signal output to the movable unit 2 or the response from the movable unit 2 to the control signal to the estimation unit 10 of the monitoring device 100. For example, a program executed by a processor or the like included in the target apparatus MA is stored in advance in a storage unit such as an EEPROM (Electrically Erasable Programmable Read-Only Memory) included in the target apparatus MA.

  Note that the control unit 1 of the target device MA may be realized by a processor included in the monitoring device 100 executing a program or the like. In this case, the processor included in the monitoring device 100 outputs a control signal for controlling the operation to the movable unit 2 via the interface unit included in the monitoring device 100. Further, the processor of the monitoring device 100 receives a response to the output control signal from the movable unit 2 via the interface unit included in the monitoring device 100. The processor of the monitoring device 100 outputs the control signal output to the movable unit 2 or the response from the movable unit 2 to the control signal to the estimation unit 10.

  The movable unit 2 includes a servo motor and the like that operate according to control by the control unit 1 and a bearing that transmits power from the servo motor and the like.

  The measuring unit 3 is, for example, a vibration sensor that measures vibration generated in the target device MA when the movable unit 2 operates under the control of the control unit 1. The measurement unit 3 converts the measured vibration into an electrical signal, and outputs the converted electrical signal to the detection unit 20 of the monitoring device 100 as data indicating the vibration in the target device MA. Note that the measurement unit 3 is not limited to a vibration sensor, and may be an acceleration sensor or an electronic gyro sensor that can detect vibration. Further, the magnitude of the vibration of the target device MA generated by the operation of the movable unit 2 varies depending on the position in the target device MA. For this reason, it is preferable that the measurement unit 3 is installed at a position where the vibration of the target device MA generated by the operation of the movable unit 2 is larger than other positions in the target device MA.

  For example, the output device 4 includes one or more of a display device including a display such as an organic EL (Organic Electro-Luminescence) or liquid crystal, a speaker, and a light. The output device 4 may be provided inside the monitoring device 100.

  For example, the estimation unit 10 transmits the vibration generated by the operation of the movable unit 2 included in the target device MA to the target device MA based on the characteristic that the target device MA transmits and the operation content of each movable unit 2. Estimate the characteristics of the vibration that occurs.

  Here, the characteristic that the target device MA transmits the vibration generated by the operation of the movable unit 2 is, for example, a transfer function that receives the operation of the movable unit 2 and outputs the vibration generated in the target device MA. . Further, the characteristic that the target device MA transmits the vibration generated by the operation of the movable unit 2 is a characteristic unique to the movable unit 2, and is evaluated, for example, when the target device MA is developed, and is included in the monitoring device 100. It is stored in advance in a storage unit such as an EEPROM.

  The operation content of the movable unit 2 includes, for example, a control signal output from the control unit 1 to the movable unit 2, a response from the movable unit 2 to the control signal, and the like. The control signal to the movable unit 2 output by the control unit 1 includes, for example, the amount of change in the position and posture of the movable unit 2 such as “rotate from an angle of 10 degrees to an angle of 30 degrees in 500 milliseconds” Instructions to specify speed etc. are included. The response from the movable unit 2 to the control signal of the control unit 1 includes, for example, data indicating the position such as the current angle. Then, the control unit 1 receives at least one of the control signal and the response from the movable unit 2 as the operation content of the movable unit 2, and the received operation content is estimated via the interface unit included in the monitoring device 100. Output to.

  For example, the estimation unit 10 reads, from a storage unit such as an EEPROM included in the monitoring device 100, a transfer function evaluated in advance as a characteristic that the target device MA transmits vibration. Further, the estimation unit 10 receives from the control unit 1 a control signal of the control unit 1 or a response from the movable unit 2 to the control signal as an operation content of the movable unit 2 through an interface unit included in the monitoring device 100. To do. Then, the estimation unit 10 performs a binomial operation or the like based on the read transfer function and the received operation content of the movable unit 2, thereby obtaining a characteristic of vibration generated in the target device MA due to the operation of the movable unit 2. presume.

  For example, the detection unit 20 compares the vibration characteristic of the target device MA estimated by the estimation unit 10 with the vibration property of the target device MA measured by the measurement unit 3, thereby detecting the abnormality of the movable unit 2. To detect. For example, the detection unit 20 performs frequency analysis on the data indicating the vibration of the target device MA received from the measurement unit 3, and obtains the frequency characteristic of the signal of the target device MA as an example of the vibration characteristic of the target device MA. . For example, when the vibration characteristic of the target device MA estimated by the estimation unit 10 is similar to the measured vibration characteristic of the target device MA, the detection unit 20 determines that the movable unit 2 is normal. On the other hand, when the similarity between the estimated vibration characteristic of the target device MA and the measured vibration characteristic of the target device MA is lower than a predetermined value, the detection unit 20 generates an abnormality in the movable unit 2. It is determined that As a process of comparing the estimated vibration characteristics of the target device MA with the measured vibration characteristics of the target device MA, the detection unit 20 compares the measured vibration characteristics of the target device MA with the estimated characteristics. A process of calculating a cross-correlation coefficient between them may be executed. Then, for example, when the calculated cross-correlation coefficient value is equal to or greater than a predetermined threshold, the detection unit 20 determines that the movable unit 2 is normal, and the cross-correlation coefficient value is smaller than the predetermined threshold. In this case, it is determined that an abnormality has occurred in the movable part 2.

  For example, the detection unit 20 outputs information on the movable unit 2 in which an abnormality has been detected to the output device 4 connected to the monitoring device 100. Thereby, the administrator of the monitoring device 100 can promptly respond to the movable unit 2 in which an abnormality is detected based on the information output from the output device 4.

  FIG. 2 shows an example of monitoring processing in the monitoring apparatus 100 shown in FIG. Steps S10 and S20 are executed when a processor mounted on the monitoring apparatus 100 executes a monitoring program. That is, FIG. 2 shows an embodiment of a monitoring program and a monitoring method. In this case, the estimation unit 10 and the detection unit 20 illustrated in FIG. 1 are realized by executing a monitoring program. Note that the processing shown in FIG. 2 may be executed by hardware installed in the monitoring apparatus 100. In this case, the estimation unit 10 and the detection unit 20 illustrated in FIG. 1 are realized by a circuit arranged in the monitoring device 100.

  In step S <b> 10, the estimation unit 10, as described with reference to FIG. 1, based on the characteristic that the target device MA transmits the vibration generated by the operation of the movable unit 2 and the operation content of the movable unit 2. Estimate the characteristics of vibration generated in

  Next, in step S <b> 20, as described with reference to FIG. 1, the detection unit 20 determines the vibration characteristics of the target device MA estimated by the estimation unit 10 and the vibration characteristics of the target device MA measured by the measurement unit 3. Is detected, the abnormality of the movable part 2 is detected. For example, when an abnormality is detected, information indicating the abnormality is output to the output device 4.

  Then, the monitoring process by the monitoring device 100 ends. Note that the flow shown in FIG. 2 may be repeatedly executed or may be executed at a predetermined frequency.

  As described above, in this embodiment, the estimation unit 10 determines the vibration characteristics of the target device MA based on the characteristics of the target device MA transmitting the vibration generated by the operation of the movable unit 2 and the operation content of the movable unit 2. presume. The detection unit 20 is limited to a predetermined operation by comparing the vibration characteristics of the target device MA estimated by the estimation unit 10 with the vibration characteristics of the target device MA measured by the measurement unit 3. Therefore, the abnormality of the movable part 2 can be detected.

  In addition, although the monitoring apparatus 100 is arrange | positioned outside the target apparatus MA, the monitoring apparatus 100 may be arrange | positioned in the target apparatus MA, for example.

  FIG. 3 shows another embodiment of the monitoring device. The monitoring device 100 shown in FIG. 3 operates as a control device that monitors the operation of the target device MA1 and controls the operation of the target device MA1. Among the elements of the monitoring apparatus 100 shown in FIG. 3, those having the same or similar functions as those of the monitoring apparatus 100 shown in FIG. 1 are denoted by the same reference numerals, and detailed description thereof is omitted.

  For example, the target device MA1 has N movable parts 2a (2a (1) -2a (N)) and a measurement 3. (N is a positive integer of 2 or more). The movable part 2a includes a servo motor and the like arranged in the target apparatus MA1, a bearing for transmitting power from the servo motor, and the like. In addition, the movable part 2a may be one.

  For example, the monitoring device 100 includes an estimation unit 10a, a detection unit 20a, a control unit 30, and a storage unit 40. The control unit 30 functions when a processor or the like included in the monitoring device 100 executes a program or the like, and controls operations of the movable units 2a (1) -2a (N) of the target device MA1. Note that a program executed by a processor or the like included in the monitoring apparatus 100 is stored in advance in the storage unit 40 such as an EEPROM, for example.

  For example, the control unit 30 selects any one of the movable units 2a (1) -2a (N) and outputs a control signal for operating the selected movable unit 2a (j) to the movable unit 2a (j). (J is a positive integer from 1 to N). For example, the control unit 30 outputs a control signal including an instruction for designating a change amount of the position and posture of the movable unit 2a (j), an exercise speed, and the like to the movable unit 2a (j). The control signal output from the control unit 30 is, for example, an instruction to rotate the current angle from −30 degrees to 30 degrees in 500 milliseconds, or a distance of 10 cm from the current position in a predetermined direction in 1 second. Instructions such as instructions to move are included. For example, the control unit 30 sequentially receives, from the movable unit 2a (j), responses such as the angle at which the movable unit 2a (j) is rotated and the distance moved by the control signal. The control unit 30 calculates an angular velocity and a moving speed from the received angle, distance, etc., and uses the calculated angular velocity, moving speed, etc. as information indicating the operation content of the movable unit 2a (j) along with the received angle, distance, etc. It outputs to the estimation part 10a. In the following description, the calculated angular speed, moving speed, etc. of the movable part 2a (j) are represented as an action function Aj (t) with the time t as a variable.

  The storage unit 40 is a storage area to which information indicating a transfer function obtained by preliminarily evaluating characteristics transmitted by the target device MA1 caused by each operation of the movable units 2a (1) -2a (N) is assigned. To store. In FIG. 3, the information stored in advance in the storage area assigned to the movable part 2a (j) is a frequency indicating the characteristic that vibrations are transmitted by the target device MA1 when the movable part 2a (j) operates. It is shown as a transfer function Hj (f) that is a function of f and intensity. The transfer function Hj (f) will be described with reference to FIG.

  For example, the estimation unit 10a reads from the storage unit 40 the transfer function Hj (f) when the movable unit 2a (j) selected by the control unit 30 operates. The estimation unit 10a estimates the characteristics of vibration generated in the target device MA1 based on the motion function Aj (t) of the movable unit 2a (j) received from the control unit 30 and the read transfer function Hj (f). To do. The estimation unit 10a outputs the estimated vibration characteristics of the target device MA1 to the detection unit 20a. In FIG. 3, the vibration characteristic of the target apparatus MA1 estimated by the estimation unit 10a is shown as a function Ej (t, f) of the frequency f and the time t. The function Ej (t, f) will be described with reference to FIG. 4 together with the operation of the estimation unit 10a.

  The detection unit 20a compares the vibration characteristic of the target device MA1 estimated by the estimation unit 10a with the vibration characteristic of the target device MA1 measured by the measurement unit 3, thereby selecting the selected movable unit 2a (j ) Is detected. For example, when detecting the abnormality of the selected movable part 2a (j), the detection unit 20a outputs information indicating the abnormality to the output device 4. The operation of the detection unit 20a will be described with reference to FIG.

  FIG. 4 shows an example of the operation function Aj (t) of the movable part 2a (j), the operation of the movable part 2a (j) shown in FIG. 3, and the measured vibration characteristics of the target apparatus MA1. FIG. 4A shows, as an operation function Aj (t), a change in angular velocity with time when the movable unit 2a (j) is rotated based on a control signal from the control unit 30 when the movable unit 2a (j) is a servo motor or the like. Show. In FIG. 4A, the vertical axis indicates angular velocity (rad / second), and the horizontal axis indicates time t. When the movable unit 2a (j) performs linear motion or the like based on the control signal from the control unit 30, the operation function Aj (t) of the movable unit 2a (j) is shown as a change in the moving speed with time.

  Time t0 indicates the time when the control unit 30 outputs a control signal to the movable unit 2a (j). A time t1 indicates a time t when the movable unit 2a (j) starts to operate based on a control signal from the control unit 30 and responds to the control unit 30. A period from time t2 to time t3 indicates a steady state period in which the angular velocity of the movable portion 2a (j) is constant. Time t4 indicates the time when the operation of the movable part 2a (j) is completed. Further, as shown in FIG. 4 (a), the time t1 and the time t0 after the control unit 30 outputs the control signal to the movable unit 2a (j) until the movable unit 2a (j) starts operating. A delay time Δt, which is the difference between

  When the movable unit 2a (j) performs linear motion based on the control signal output by the control unit 30 at time t0, the period from time t1 to time t2 is the same as in the angular velocity shown in FIG. Then, the movement speed of the motion function Aj (t) increases. In the period from time t2 to time t3, the movable part 2a (j) is in a steady state in which it moves linearly at a constant movement speed, and the motion function Aj (t) has a constant movement speed. In the period from the time t3 to the time t4, since the movable part 2a (j) decelerates the moving speed toward the end of the linear motion, the moving speed of the motion function Aj (t) decreases.

  FIG. 4B shows, for example, the time change of the angle at which the movable part 2a (j) selected by the control signal from the control part 30 is rotated. In FIG. 4B, the vertical axis indicates the angle deg and the horizontal axis indicates time t. As shown in FIG. 4B, the movable portion 2a (j) rotates from the angle α0 to the angle α1 with the angular velocity operation function Aj (t) shown in FIG.

  FIG. 4C shows, for example, the case where the selected movable unit 2a (j) operates as shown in FIGS. 4A and 4B based on the control signal from the control unit 30. The example of the waveform of the vibration of object apparatus MA1 measured by the measurement part 3 is shown. The vertical axis indicates the amplitude of vibration, and the horizontal axis indicates time t. As shown in FIG. 4 (c), the measured waveform of the vibration of the target device MA1 is larger than that in other periods due to the impact accompanying the start of operation of the movable part 2a (j) in the period from time t1 to time t2. Indicates the amplitude. On the other hand, the vibration waveform in the period from the time t2 to the time t3 when the angular velocity of the movable part 2a (j) is in a steady state shows a constant amplitude over time. The vibration waveform in the period from the time t3 to the time t4 shows a change in which the amplitude is attenuated with the lapse of time because the angular velocity of the movable portion 2a (j) decreases toward the end of the operation.

  FIG. 4D shows a frequency distribution Rj (t, f) of vibration obtained by performing frequency analysis on the vibration waveform shown in FIG. For example, the detection unit 20 samples the vibration waveform of the target apparatus MA1 measured by the measurement unit 3 shown in FIG. 4C at a predetermined sampling frequency (16 kilohertz, 22 kilohertz, etc.). For example, the detection unit 20a performs frequency analysis such as FFT (Fast Fourier Transform) for each predetermined sample unit such as 1024 (also referred to as spectrogram), and the vibration frequency distribution Rj (t, f shown in FIG. ) In FIG.4 (d), a vertical axis | shaft shows the frequency f (Hz) and a horizontal axis shows the time t. In FIG. 4D, the intensity of the frequency distribution Rj (t, f) is shaded in the space indicated by the time axis t, the frequency axis f, and the intensity (black has a high intensity and white has a low intensity). Shows the distribution of the intensity of each frequency component. In the frequency distribution Rj (t, f) shown in FIG. 4D, the frequency conversion window for frequency analysis such as FFT is set larger than in the case of FIG. 4E described later. For this reason, the frequency distribution Rj (t, f) shown in FIG. 4 (d) has a greater effect of smoothing than that of FIG. 4 (e), so that the relationship with the time change of the motion function Aj (t) is known. Hateful.

  FIG. 4 (e) schematically shows, for example, a frequency distribution Rj (t, f) obtained by extracting a portion of the frequency distribution Rj (t, f) shown in FIG. Show. The vertical axis indicates the frequency f of vibration, and the horizontal axis indicates time t. In the frequency distribution Rj (t, f) shown in FIG. 4 (e), the frequency conversion window for frequency analysis is set smaller than in the case of FIG. 4 (d). Therefore, the frequency distribution Rj (t, f) shown in FIG. 4 (e) is less influenced by smoothing by frequency analysis than the case of FIG. 4 (d), and the intensity according to the time change of the operation function Aj (t). A transition of the frequency f of the distribution appears. For example, in the period from time t1 to time t2, the frequency distribution Rj (t, f) shown in FIG. 4 (e) has strong strength because the change in angular velocity is large due to the start of operation of the movable part 2a (j). The intensity distribution of the frequency band transitions from a low frequency to a high frequency with respect to time. On the other hand, in the period from time t2 to time t3 when the operation of the movable portion 2a (j) is in a steady state, the frequency distribution Rj (t, f) has a strong frequency band that is substantially constant with respect to time (frequency No distribution transition). In the period from time t3 to time t4, the angular velocity of the movable portion 2a (j) decreases toward the end of the operation, so that the intensity distribution of the measured vibration frequency distribution Rj (t, f) Transition from high frequency to low frequency.

  By the way, when the movable part 2a (j) is operating normally, the measured frequency distribution Rj (t, f) of the vibration of the target device MA1 is the operation of the movable part 2a (j) every time t. It is calculated by superimposing the function Aj (t) and the transfer function Hj (f) of the frequency f corresponding to the angular velocity of the motion function Aj (t) and the intensity. That is, the frequency distribution Rj (t, f) represents the relationship between the frequency f of the vibration generated by the operation of the movable part 2a (j) at every time t and its intensity, and is constant if the angular velocity is the same. At the frequency f, ideally the same intensity distribution is shown.

  Hereinafter, a method of acquiring the transfer function Hj (f) when the movable unit 2a (j) operates by using the functions of the detection unit 20a and the control unit 30 included in the monitoring device 100 will be described. In the following description, the transfer function Hj (f) when the movable part 2a (j) operates is referred to as the transfer function Hj (f) of the movable part 2a (j).

  In order to obtain the transfer function Hj (f) of the movable part 2a (j), the control unit 30 performs a predetermined operation in the development stage, the shipping stage, etc. of the target apparatus MA1 and the monitoring apparatus 100, for example. 2a (j) is performed a predetermined number of times (50 times, etc.). For example, the detection unit 20a samples the vibration waveform of the target device MA1 measured by the measurement unit 3 every time the movable unit 2a (j) operates at a predetermined sampling frequency, and sequentially stores the vibration data in the storage unit 40 as vibration data. To remember. For example, after the predetermined number of operations by the movable portion 2a (j) are completed, the detection unit 20a performs averaging of the stored predetermined number of vibration data. For example, the detection unit 20a compares the vibration data of the period from time t2 to time t3 in which the operation of the movable unit 2a (j) is in a steady state and the frequency distribution is constant in the time axis direction. Then, frequency analysis is performed in a predetermined sample unit. The detection unit 20a acquires the frequency distribution obtained from the frequency analysis as the transfer function Hj (f) of the movable unit 2a (j). Then, the detection unit 20a stores the acquired transfer function Hj (f) of the movable unit 2a (j) in the storage unit 40.

  The predetermined operations to be performed by the movable part 2a (j) include the start position and end position of the operation, a predetermined angular velocity (for example, 10 radians per second), and a predetermined movement speed (for example, 5 centimeters per second). Etc.) is preset for each movable part 2a. That is, the obtained transfer function Hj (f) is a transfer function when the movable part 2a (j) operates at a predetermined angular velocity or a predetermined movement speed. Therefore, the detection unit 20a corresponds to the acquired transfer function Hj (f) with a predetermined angular velocity or a predetermined moving speed indicating an operation performed by the movable unit 2a (j) when acquiring the transfer function Hj (f). You may add and memorize | store in the memory | storage part 40.

  In addition, when adding and averaging a plurality of vibration data, the detection unit 20a obtains a position where the cross-correlation value obtained by performing a cross-correlation process between each of the plurality of vibration data is maximized. The detection unit 20a preferably corrects each phase shift between the plurality of vibration data with the obtained position as a reference, and averages the corrected vibration data.

  Furthermore, the control unit 30 outputs a delay time Δt from the output of the control signal for causing the movable unit 2a (j) to perform a predetermined operation to the reception of the first response from the movable unit 2a (j) as the time t1. It is obtained from the difference from time t0. The control unit 30 may store the obtained delay time Δt in the storage unit 40 in association with the transfer function Hj (f) of the movable unit 2a (j) acquired by the detection unit 20a.

  The transfer function Hj (f) obtained as described above is used by the estimation unit 10a for the process of estimating the characteristics of vibration generated in the target apparatus MA1 when the movable unit 2a (j) operates. .

  For example, the estimation unit 10a reads the transfer function Hj (f) from the storage unit 40 as a characteristic that the target device MA1 transmits the vibration generated by the operation of the movable unit 2a (j) selected by the control unit 30. The estimation unit 10a performs a superposition calculation of the motion function Aj (t) of the movable unit 2a (j) received from the control unit 30 and the read transfer function Hj (f), and vibrations generated in the target device MA1. The expected value Ej (t, f) of the frequency distribution of vibration, which is the characteristic of In estimating the expected value Ej (t, f) of the frequency distribution, the estimating unit 10a delays Δt from when the control unit 30 outputs a control signal to the movable unit 2a (j) until receiving a response. May be considered. Thereby, the detection unit 20a can compare the measured frequency distribution Rj (t, f) of the vibration of the target device MA1 with the estimated expected value Ej (t, f) of the frequency distribution in real time. it can.

  The angular velocity and the moving velocity indicated by the motion function Aj (t) of the movable portion 2a (j) received by the estimation portion 10a from the control portion 30 are obtained when the transfer function Hj (f) of the movable portion 2a (j) is acquired. It may be different from the predetermined angular velocity or the predetermined moving speed. Therefore, the estimation unit 10a, for example, the angular velocity and movement speed indicated by the motion function Aj (t) received from the control unit 30, and the predetermined angular velocity and predetermined movement speed when the transfer function Hj (f) is acquired. It is preferable to normalize the motion function Aj (t) based on the ratio of. By using the normalized motion function Aj (t), the estimation unit 10a can obtain an angular velocity different from the predetermined angular velocity or the predetermined moving velocity when the movable portion 2a (j) acquires the transfer function Hj (f) The expected value Ej (t, f) of the frequency distribution when operating at the moving speed can be estimated.

  For example, the detection unit 20a performs a cross-correlation between the vibration frequency distribution Rj (t, f) measured by the measurement unit 3 and the expected value Ej (t, f) of the frequency distribution estimated by the estimation unit 10a. Based on the result obtained by performing the process, the abnormality of the movable portion 2a (j) is detected. For example, the detection unit 20a performs a cross-correlation process with respect to the amplitude in the frequency axis direction between the frequency distribution Rj (t, f) and the expected value Ej (t, f) at every predetermined sample unit time interval. Find the cross correlation coefficient. The detection unit 20a compares the maximum value of the cross-correlation coefficient of the movable unit 2a (j) obtained by the cross-correlation process with a predetermined threshold (for example, 0.5) in a predetermined sample unit time. . The detection unit 20a obtains the number of samples in a predetermined sample unit in which the maximum value of the cross-correlation coefficient of the movable unit 2a (j) is determined to be equal to or greater than a predetermined threshold. The detection unit 20a performs the cross-correlation process of the amplitude in the frequency axis direction between the frequency distribution Rj (t, f) and the expected value Ej (t, f), for example, in the range of 100 to 1000 hertz. It is preferable to carry out with a width of about ± 50 hertz. By giving a width of about ± 50 Hz, the cross-correlation process in the detection unit 20a can absorb frequency errors due to noise or the like superimposed on the frequency distribution Rj (t, f).

  FIG. 5 shows an example of monitoring processing in the monitoring apparatus 100 shown in FIG. Steps S100, S110, S120, S130, S140, S150, S160, S170, S180, and S190 are executed when a processor mounted on the monitoring apparatus 100 executes a monitoring program. That is, FIG. 5 shows another embodiment of the monitoring program and the monitoring method. In this case, the estimation unit 10a, the detection unit 20a, and the control unit 30 illustrated in FIG. 3 are realized by executing the monitoring program. Note that the processing shown in FIG. 5 may be executed by hardware installed in the monitoring apparatus 100. In this case, the estimation unit 10 a, the detection unit 20 a, and the control unit 30 illustrated in FIG. 3 are realized by a circuit arranged in the monitoring device 100.

  In step S100, the control unit 30 outputs a control signal to the selected movable unit 2a (j) as described in FIG.

  Next, in step S110, the control unit 30 obtains an operation function Aj (t) of the movable unit 2a (j) based on responses sequentially received from the movable unit 2a (j) with respect to the output control signal.

  Next, in step S120, as described with reference to FIGS. 3 and 4, the detection unit 20a samples the electrical signal indicating the vibration of the target device MA1 measured by the measurement unit 3 at a predetermined sampling frequency, and vibrates. Get as data.

  Next, in step S130, for example, the detection unit 20a performs frequency analysis such as FFT on the acquired vibration data for each predetermined sample unit, thereby measuring the frequency distribution Rj of the vibration of the measured target device MA1. Calculate (t, f).

  Next, in step S140, the estimation unit 10a performs a superposition calculation of the motion function Aj (t) obtained in step S110 and the transfer function Hj (f) of the movable unit 2a (j) read from the storage unit 40. By executing, the expected value Ej (t, f) of the frequency distribution is estimated. The expected value Ej (t, f) is a transfer function Hj (f) that is a function of frequency and intensity, a normalized angular velocity normalized at the same angular velocity as when the transfer function Hj (f) was acquired, and a time t. The time t, the frequency obtained by multiplying the frequency axis of the transfer function Hj (f) by the normalized angular velocity of the action function Aj (t) from the action function Aj (t), which is a function, at each time t. It is an intensity distribution (spectrogram) for each frequency.

  Next, in step S150, the detection unit 20a is estimated as the frequency distribution Rj (t, f) of the vibration of the target apparatus MA1 at time intervals of a predetermined sample unit, as described with reference to FIGS. A cross-correlation process is performed with the expected value Ej (t, f).

  Next, in step S160, the detection unit 20a determines whether or not the obtained maximum value of the cross-correlation coefficient of the movable unit 2a (j) is equal to or greater than a predetermined threshold, as described with reference to FIGS. To do. The detection unit 20a obtains the number of samples in a predetermined sample unit in which the maximum value of the cross correlation coefficient is equal to or greater than a predetermined threshold.

  Next, in step S170, as described with reference to FIG. 3 and FIG. 4, the detection unit 20a includes the ratio of the number of samples in which the maximum value of the cross-correlation coefficient is equal to or greater than a predetermined threshold in the movable unit 2a (j). It is compared with a predetermined ratio to determine whether or not an abnormality has occurred in each movable part 2a (j). When the ratio of the number of samples in which the maximum value of the cross-correlation coefficient is equal to or greater than a predetermined threshold is smaller than the predetermined ratio (YES), the detection unit 20a determines that an abnormality has occurred in the movable unit 2a (j). The process proceeds to step S180. On the other hand, when the ratio of the number of samples in which the maximum value of the cross-correlation coefficient is equal to or greater than a predetermined threshold is greater than or equal to a predetermined ratio (NO), the detection unit 20a indicates that the movable unit 2a (j) is operating normally. Determination is made, and the process proceeds to step S190.

  In step S180, the detection unit 20a outputs an alarm including information on the movable unit 2a (j) in which an abnormality is detected in the output device 4.

  Next, in step S190, the monitoring apparatus 100 determines whether or not an end instruction has been received via an input device such as a keyboard or a touch panel included in the monitoring apparatus 100, for example. When the monitoring apparatus 100 receives an end instruction (YES), the series of processing ends. On the other hand, if the monitoring apparatus 100 has not received an end instruction (NO), the process proceeds to step S100.

  As described above, in this embodiment, the storage unit 40 uses the transfer function H1 (f) − as information indicating the characteristics of the target device MA1 to transmit the vibration generated by the operations of the movable units 2a (1) -2a (N). HN (t) is stored in advance. The estimation unit 10a estimates the expected value Ej (t, f) of the frequency distribution of the vibration of the target apparatus MA1 from the controlled operation function Aj (t) and transfer function Hj (f) of the movable unit 2a (j). . The detection unit 20a obtains a cross-correlation coefficient obtained by performing a cross-correlation process between the measured frequency distribution Rj (t, f) and the estimated expected value Ej (t, f) and a predetermined threshold value. Compare. Thereby, the detection unit 20a can detect an abnormality of the movable unit 2a (j) without being limited to a predetermined operation.

  In addition, the detection unit 20a detects an abnormality of the selected movable unit 2a (j) based on the cross-correlation process between the measured frequency distribution Rj (t, f) and the estimated expected value Ej (t, f). Is detected. For this reason, the abnormality of the selected movable part 2a (j) can be detected using the measurement part 3 common to the movable parts 2a (1) -2a (N). Thereby, compared with the case where the measurement part 3 is arrange | positioned to each movable part 2a, the number of the measurement parts 3 can be decreased, and cost reduction can be aimed at.

  Further, the detection unit 20a outputs an alarm including information on the movable unit 2a (j) in which an abnormality is detected in the output device 4. Thereby, it is possible to prompt the administrator of the monitoring apparatus 100 to respond to the movable part 2a (j) in which the abnormality is detected, and to quickly repair or replace the movable part 2a (j). it can.

  In FIG. 3, one measuring unit 3 is arranged in the target device MA1, but the present invention is not limited to this. For example, the number of measurement units 3 arranged in the target apparatus MA1 may be two or more and less than N, which is the number of movable units 2a (1) -2a (N). In this case, the measurement part 3 is arrange | positioned in the position which mutually separated in object apparatus MA1. For example, among the plurality of measurement units 3, a measurement unit 3 capable of measuring vibrations generated by the operation of each movable unit 2a more efficiently than the other measurement units 3 is preset for each movable unit 2a, and the detection unit 20a Uses vibration data of the target device MA1 measured by the measuring unit 3 set for each movable unit 2a. In this case, when acquiring the transfer function of each movable part 2a, it is preferable that the detection part 20a uses, for example, vibration data of the target device MA1 measured by the measurement part 3 set for each movable part 2a.

  Note that the monitoring process in the monitoring apparatus 100 is not limited to the order of the processes illustrated in FIG. 5. For example, the processes in steps S110 and S140 and the processes in steps S120 and S130 may be performed in parallel. .

  The monitoring device 100 operates as a control device that monitors the operation of the target device MA1 and controls the operation of the target device MA1, but is not limited thereto. For example, a processor or the like included in the target device MA1 functions as a control unit, the processor or the like of the target device MA1 outputs a control signal to the movable unit 2a (j), and the movable unit 2a (j) corresponding to the output control signal. A response from may be received. Then, the processor or the like of the target device MA1 may output the received response of the movable unit 2a (1) -2a (N) to the control unit 30a. Further, instead of the control unit 30 of the monitoring device 100, the processor or the like of the target device MA1 obtains the operation function Aj (t) based on the response from the movable unit 2a (j) and outputs it to the estimation unit 10a. Good.

  Further, although the monitoring device 100 is arranged outside the target device MA1, for example, the monitoring device 100 may be arranged in the target device MA1.

  In addition, in order to acquire the transfer function Hj (f) of each movable part 2a, the operation | movement which the control part 30 performs each movable part 2a is not limited to single operation decided beforehand, For example, a movable part A plurality of patterns of operations determined in advance for each 2a may be used. And the control part 30 is based on the frequency distribution measured by the measurement part 3 when performing the operation | movement of several predetermined patterns for every movable part 2a, respectively, and transfer function Hj ( f) may be obtained, and the obtained transfer function Hj (f) may be stored in the storage unit 40. For example, a plurality of predetermined patterns of movement include a plurality of predetermined angular velocities (e.g., 8 radians per second, 10 radians per second, etc.) and a plurality of movable ranges (-30 degrees to -10 degrees, -10 degrees to 10 degrees). 10 degrees to 30 degrees). In addition, the movement of a plurality of patterns includes a plurality of predetermined moving speeds (such as 5 centimeters per second) and a plurality of movable ranges (from 1 centimeter to 7 centimeters in the gravitational direction or in a horizontal direction perpendicular to the gravitational direction. 3 centimeters to 9 centimeters).

  By obtaining a plurality of transfer functions Hj (f) for each movable part 2a, the estimation part 10a can transfer the transfer function Hj (f) corresponding to the action of the movable part 2a (j) indicated by the action function Aj (t). ), The expected value Ej (t, f) can be calculated in consideration of the effects of friction and gravity. Thereby, compared with the case where the single transfer function Hj (f) acquired by executing one operation for each movable part 2a is used, the monitoring apparatus 100 detects the abnormality of the movable part 2a (j) more accurately. Can be done well.

  Note that the detection unit 20a performs a cross-correlation process between the measured frequency distribution Rj (t, f) and the estimated expected value Ej (t, f) during the entire operation period of the movable unit 2a (j). Although it is determined whether or not the maximum value of the obtained cross-correlation coefficient is equal to or greater than a predetermined threshold value, the present invention is not limited to this. For example, the detection unit 20a determines whether or not the maximum value of the calculated cross-correlation coefficient is equal to or greater than a predetermined threshold during a period in which the operation of the movable unit 2a (j) is in a steady state, and a period other than the steady state The determination at may be omitted. Thereby, it is possible to avoid the influence of the impact at the start of the operation of the movable part 2a (j), the noise generated at the end of the operation, etc. It can be avoided. Moreover, since the determination period can be shortened by omitting the determination in a period other than the steady state, the monitoring process in the monitoring apparatus 100 can be speeded up.

  The detection unit 20a determines whether or not the ratio of the number of samples determined that the maximum value of the cross-correlation coefficient of the movable unit 2a (j) is equal to or greater than a predetermined threshold is smaller than the predetermined rate, and the movable unit 2a Although the abnormality of (j) was detected, it is not limited to this. For example, the detection unit 20a determines that there is a possibility of abnormality when the maximum value of the cross-correlation coefficient is determined to be smaller than a predetermined threshold for the movable unit 2a (j). A determination result indicating that there is a possibility of abnormality may be output to the output device 4. In this case, after the monitoring process shown in FIG. 5 is completed, for example, the control unit 30 can operate the movable unit 2a (j) with respect to the movable unit 2a (j) that is determined to have an abnormality. A control signal for inspection or the like for operating the movable part 2a (j) may be output within such a range. And the monitoring apparatus 100 is more accurate than the monitoring process shown in FIG. 5 based on the frequency distribution measured by the measurement part 3 when the movable part 2a (j) operated according to the control signal for inspection. An inspection may be performed. Thereby, the monitoring device 100 can acquire detailed information about the abnormality that has occurred in the movable portion 2a (j), and the administrator of the monitoring device 100 can move based on the detailed information that is output to the output device 4. Appropriate measures can be taken for the abnormality of the part 2a (j). However, it is preferable that the monitoring device 100 has a function of confirming whether or not there is a person around when performing a highly accurate inspection, for example, when the target device MA1 is a robot that communicates with a person.

  FIG. 6 shows another embodiment of the monitoring device. The monitoring device 100 illustrated in FIG. 6 operates as a control device that monitors the operation of the target device MA1 and controls the operation of the target device MA1 in the same manner as the monitoring device 100 illustrated in FIG. Among the elements of the monitoring apparatus 100 shown in FIG. 6, those having the same or similar functions as those of the monitoring apparatus 100 shown in FIG.

  The configuration of the target device MA1 is the same as or similar to the target device MA1 shown in FIG. The monitoring device 100 includes N estimation units 10b (1) -10b (N) corresponding to the movable units 2a (1) -2a (N), a detection unit 20b, a control unit 30a, and a storage unit 40.

  For example, the control unit 30a controls the operations of the movable units 2a (1) -2a (N) of the target device MA1 in parallel. For example, the control unit 30a outputs a control signal to each of the movable units 2a (1) -2a (N) and operates each of the movable units 2a (1) -2a (N). In addition, the control unit 30a receives a response indicating a rotated angle, a moved distance, and the like with respect to the output control signal from each of the movable units 2a (1) -2a (N) sequentially or in parallel. Based on the received response, the control unit 30a calculates the operation function A1 (t) -AN (t) of the movable unit 2a (1) -2a (N). The control unit 30a outputs the calculated operation function A1 (t) -AN (t) to the estimation units 10b (1) -10b (N) together with the received angle, distance, and the like.

  For example, the estimation unit 10b (1) -10b (N) reads the transfer function H1 (f) -HN (f) of the movable unit 2a (1) -2a (N) allocated in advance from the storage unit 40, respectively. For example, the estimation unit 10b (j) reads the transfer function Hj (f) of the movable unit 2a (j), and overlaps the read transfer function Hj (f) with the motion function Aj (t) received from the control unit 30a. Perform the alignment operation. The estimation unit 10b (j) estimates the expected value Ej (t, f) of the frequency distribution that is the characteristic of the vibration generated in the target device MA1 by the executed overlay calculation. The estimation unit 10b (1) -10b (N) outputs the estimated expected frequency distribution value E1 (t, f) -EN (t, f) to the detection unit 20b. Note that the estimation units 10b (1) -10b (N) operate in parallel.

  In addition, when the estimation unit 10b (1) -10b (N) estimates the expected value E1 (t, f) -EN (t, f), the control unit 30a outputs a control signal to each movable unit 2a. The delay time Δt of each movable portion 2a from when the response is received to when the response is received may be taken into consideration.

  In addition, the estimation unit 10b (1) -10b (N), for example, transfers the operation functions A1 (t) -AN (t) of the movable units 2a (1) -2a (N) received from the control unit 30a. You may normalize with the predetermined | prescribed angular velocity used at the time of acquisition, predetermined | prescribed moving speed, etc. As a result, the estimation unit 10b-10b (N) performs frequency distribution even when each of the movable units 2a (1) -2a (N) operates at a predetermined angular velocity or an angular velocity or movement speed different from the predetermined movement velocity. Expected value E1 (t, f) -EN (t, f) can be estimated with high accuracy.

  For example, the detection unit 20b obtains the expected value E1 (t, f) -EN (t, f) of the frequency distribution, which is the characteristic of the vibration generated in the estimated target device MA1, as the estimation unit 10b (1) -10b ( N), respectively. The detection unit 20b compares the expected value E1 (t, f) -EN (t, f) of the estimated frequency distribution with the measured vibration characteristic of the target device MA1, thereby moving the movable unit 2a (1 ) -2a (N) abnormality is detected. The operation of the detection unit 20b will be described with reference to FIGS.

  By the way, when the control unit 30a operates the M movable units 2a among the N movable units 2a, the vibrations of the target device MA1 measured by the measurement unit 3 are respectively operated by the M movable units 2a. This is equivalent to a vibration obtained by superimposing vibrations generated in the target device MA1. That is, the frequency distribution of the vibration of the target device MA1 measured by the measurement unit 3 is expressed as Expression (1). M is a positive integer less than or equal to N.

  In Expression (1), “TR (t, f)” indicates a frequency distribution of vibration of the target device MA1 measured by the measurement unit 3 when the M movable units 2a are operated. Hereinafter, the frequency distribution TR (t, f) is a sum of frequency distributions of vibrations when the M movable parts 2a are operated. In the following, the monitoring operation in the monitoring device 100 when there are three movable parts 2a controlled by the control unit 30a will be described. However, in the monitoring apparatus 100 when a plurality of movable parts 2a other than three are controlled. The same applies to the monitoring operation, and detailed description thereof is omitted.

  FIG. 7 shows an example of the operation function and the vibration frequency distribution of the target apparatus MA1 when three of the movable parts 2a shown in FIG. 6 are operated. FIG. 7 shows the characteristics of vibration measured by the target apparatus MA1. FIG. 7A shows, for example, the temporal changes in the angular velocities of the movable parts 2a (1), 2a (2), and 2a (3) as motion functions A1 (t), A2 (t), and A3 (t). . In FIG. 7A, the vertical axis indicates angular velocity (rad / sec), and the horizontal axis indicates time. The solid curve shown in FIG. 7 (a) indicates the motion function A1 (t) of the movable portion 2a (1), and the dotted curve indicates the motion function A2 (t) of the movable portion 2a (2). These curves respectively show the operation function A3 (t) of the movable part 2a (3). As shown in FIG. 7A, the movable portion 2a (1) operates at the highest angular velocity, and the movable portion 2a (2) operates at the lowest angular velocity.

  In addition, times t10, t11, and t12 indicate times at which the movable units 2a (1), 2a (2), and 2a (3) start to operate based on a control signal from the control unit 30a. In addition, times t13, t15, and t14 indicate times when the movable units 2a (1), 2a (2), and 2a (3) have finished their operations based on the control signal from the control unit 30a. That is, as shown in FIG. 7A, the movable part 2a (1) operates in the period from time t10 to time t11, and in the period from time t11 to time t12, the movable part 2a (1) and the movable part 2a. (2) operates. Further, the movable portions 2a (1), 2a (2), and 2a (3) operate during the period from the time t12 to the time t13, and during the period from the time t13 to the time t14, the movable portion 2a (2) and the movable portion 2a. (3) operates, and the movable portion 2a (2) operates during the period from time t14 to time t15.

  7 (b) to 7 (d) show the movable parts 2a (1), 2a (2), and 2a (3) as shown in FIG. 7 (a) as in the case of FIG. 4 (e). The frequency distribution of the vibration which generate | occur | produces in object apparatus MA1 at the time of operate | moving independently by operation function A1 (t), A2 (t), A3 (t) is shown typically. That is, FIG. 7B shows a frequency distribution R1 (t, f) of vibration of the target apparatus MA1 when the movable part 2a (1) operates. FIG. 7 (c) shows a frequency distribution R2 (t, f) of vibration of the target apparatus MA1 when the movable part 2a (2) is operated, and FIG. 7 (d) is an operation of the movable part 2a (3). Shows the frequency distribution R3 (t, f) of the vibration of the target device MA1. 7B to 7D, the vertical axis indicates the frequency f (Hz), and the horizontal axis indicates time t.

  The frequency distributions R1 (t, f), R2 (t, f), R3 (t) of the movable parts 2a (1), 2a (2), 2a (3) shown in FIG. 7 (b) to FIG. 7 (d) , F) includes the distribution at the start of the operation and the distribution of each period at the steady state and at the end of the operation, as described in FIG. As shown in FIGS. 7B to 7D, the movable parts 2a (1), 2a (2), and 2a (3) have different frequency distributions. For example, in the frequency distribution R1 (t, f) of the movable part 2a (1) shown in FIG. 7B, as in FIG. 4E, when the operation starts, the operation of the movable part 2a (1) is started. Due to the rapid change in angular velocity, the intensity distribution of the frequency band having strong intensity transitions from a low frequency to a high frequency with respect to time. Further, in the frequency distribution R1 (t, f), the frequency band having a strong intensity is substantially constant with respect to time during the period in which the operation of the movable part 2a (1) is in a steady state (there is no frequency distribution transition). ). In the period in which the movable portion 2a (1) decreases the angular velocity toward the end of the operation, the intensity distribution of the frequency distribution R1 (t, f) transitions from a high frequency to a low frequency with respect to time.

  In the frequency distribution R2 (t, f) of the movable part 2a (2) shown in FIG. 7 (c), the operation of the movable part 2a (2) starts at the start of the operation, as in the case of the movable part 2a (1). Due to the rapid change in angular velocity associated with, the intensity distribution in the frequency band having strong intensity transitions from a low frequency to a high frequency with respect to time. On the other hand, in the period of the steady state of the movable part 2a (2), the frequency distribution R2 (t, f) has a frequency band with a strong intensity substantially constant with respect to time (no frequency band transition). Note that the frequency distribution R2 (t, f) differs from the frequency distribution R1 (t, f) in a frequency band having a strong intensity due to a difference in angular velocity and transfer function. In a period in which the movable portion 2a (2) decreases the angular velocity toward the end of the operation, the intensity distribution of the frequency distribution R2 (t, f) transitions from a high frequency to a low frequency with respect to time.

  Further, the frequency distribution R3 (t, f) of the movable part 2a (3) shown in FIG. 7 (d) is similar to that of FIG. 7 (b) and FIG. Due to the rapid change in angular velocity at the start of the operation 3), the intensity distribution of the frequency band having a strong intensity transitions from a low frequency to a high frequency with respect to time. On the other hand, in the steady state period of the movable part 2a (3), the frequency distribution R3 (t, f) has a frequency band with strong intensity substantially constant with respect to time (no frequency band transition). Note that the frequency distribution R3 (t, f) differs from the frequency distributions R1 (t, f) and R2 (t, f) in a frequency band having a strong intensity due to a difference in angular velocity and transfer function. In the period in which the movable portion 2a (3) decreases the angular velocity toward the end of the operation, the intensity distribution of the frequency distribution R3 (t, f) transitions from a high frequency to a low frequency with respect to time.

  FIG. 7 (e) shows frequency distributions R1 (t, f), R2 (t, t) of the movable parts 2a (1), 2a (2), 2a (3) shown in FIG. 7 (b) to FIG. 7 (d). A frequency distribution TR (t, f) obtained by summing up f) and R3 (t, f) based on the equation (1) is shown.

  As shown in FIG. 7E, since the movable part 2a (1) operates and the other movable part 2a does not operate during the period from the time t10 to the time t11, the total frequency distribution TR (t, f) Is equivalent to the frequency distribution R1 (t, f) of the movable part 2a (1). In the period from time t11 to time t12, the movable part 2a (1) and the movable part 2a (2) operate, and the other movable part 2a does not operate. Therefore, the total frequency distribution TR (t, f) is This is equivalent to a distribution obtained by adding the frequency distribution R1 (t, f) and the frequency distribution R2 (t, f). In the period from time t12 to time t13, the movable portions 2a (1), 2a (2), and 2a (3) operate, and the other movable portions 2a (j) do not operate. Therefore, the total frequency distribution TR (t , F) is equivalent to a distribution obtained by adding the frequency distributions R1 (t, f), R2 (t, f), and R3 (t, f). In the period from time t13 to time t14, the movable part 2a (2) and the movable part 2a (3) operate, and the other movable part 2a (j) does not operate. Therefore, the total frequency distribution TR (t, f ) Is equivalent to a distribution obtained by adding the frequency distribution R2 (t, f) and the frequency distribution R3 (t, f). In the period from time t14 to time t15, the movable part 2a (2) operates and the other movable parts 2a (j) do not operate. Therefore, the total frequency distribution TR (t, f) is the movable part 2a (2 ) Frequency distribution R2 (t, f).

The detection unit 20b uses the expression (2) and the frequencies of the movable units 2a (1), 2a (2), and 2a (3) estimated by the estimation units 10b (1), 10b (2), and 10b (3). A total expected value TE (t, f) is obtained from the expected values E1 (t, f), E2 (t, f), and E3 (t, f) of the distribution.
TE (t, f) = β1 × E1 (t, f) + β2 × E2 (t, f) + β3 × E3 (t, f) + nr (f) (2)
Here, “TR (t, f)” indicates the sum of expected values E1 (t, f), E2 (t, f), E3 (t, f), and coefficients β1, β2, and β3 indicate weighting coefficients. . “Nr (f)” indicates a frequency distribution of noise components and the like included in the total frequency distribution TR (t, f).

  Then, the detection unit 20b compares the total expected value TE (t, f) shown in Equation (2) with the total frequency distribution TR (t, f) shown in Equation (1). The detection unit 20b detects an abnormality in each of the movable units 2a (1), 2a (2), and 2a (3) based on the comparison result.

  For example, the detection unit 20b uses the least square method for the amplitude in the frequency axis direction between the total frequency distribution TR (t, f) and the total expected value TE (t, f) illustrated in FIG. The processing is performed at a predetermined sample unit time interval such as 1024 to obtain weight coefficients β1, β2, and β3. The detection unit 20b detects the values of the weighting factors β1, β2, and β3 obtained in predetermined sample units and predetermined threshold values (for example, during the operation period of the movable units 2a (1), 2a (2), and 2a (3)). , 0.5 etc.). In the operation period of each movable part 2a, the detection unit 20b determines the number of predetermined sample units in which the values of the respective weight coefficients β1, β2, and β3 are determined to be equal to or greater than a predetermined threshold as the movable parts 2a (1), 2a ( Obtained as the number of samples for each of 2) and 2a (3).

  For example, the detection unit 20b determines that the ratio of the number of samples determined to be equal to or greater than a predetermined threshold for each of the weighting factors β1, β2, and β3 is a predetermined ratio of the total number of predetermined sample units included in the operation period of each movable unit 2a. It is determined whether or not (for example, 70%, 80%, etc.) or more. For example, the detection unit 20b determines that the movable unit 2a (1) is operating normally when the ratio of the number of samples from which the weight coefficient β1 equal to or greater than a predetermined threshold is obtained is equal to or greater than a predetermined ratio. On the other hand, the detection unit 20b determines that an abnormality has occurred in the movable unit 2a (1) when the ratio of the number of samples from which the weighting coefficient β1 equal to or greater than the predetermined threshold is smaller than the predetermined ratio. In addition, the detection unit 20b determines that the movable unit 2a (2) is operating normally when the ratio of the number of samples from which the weighting coefficient β2 equal to or greater than the predetermined threshold is greater than the predetermined ratio, and determines the number of samples. Is smaller than the predetermined ratio, it is determined that an abnormality has occurred in the movable part 2a (2). In addition, the detection unit 20b determines that the movable unit 2a (3) is operating normally when the ratio of the number of samples from which the weighting coefficient β3 equal to or greater than the predetermined threshold is greater than the predetermined ratio, and determines the number of samples. Is smaller than the predetermined ratio, it is determined that an abnormality has occurred in the movable part 2a (3).

  FIG. 8 shows an example of temporal changes in the weighting factors β1, β2, and β3 obtained by the detection unit 20b shown in FIG. FIG. 8A shows the total frequency distribution TR (t, f) as in FIG. 7E. FIGS. 8B to 8D show temporal changes of the weighting coefficients β1, β2, and β3 obtained by the detection unit 20b. 8 (b) to 8 (d), the vertical axis indicates the values of the weighting factors β1, β2, and β3, and the horizontal axis indicates time t. “Γ” represents a predetermined threshold value.

  As shown in FIG. 8B, in the period from time t10 to time t13 when the movable part 2a (1) is operating, the weighting coefficient β1 shows a value equal to or greater than a predetermined threshold value γ. It is determined that the movable part 2a (1) is operating normally. Further, as shown in FIG. 8C, since the weighting coefficient β2 shows a value equal to or larger than a predetermined threshold γ during the period from time t11 to time t15 when the movable part 2a (2) is operating, the detection unit 20b determines that the movable part 2a (2) is operating normally. Further, as shown in FIG. 8 (d), the weighting coefficient β1 shows a value equal to or larger than a predetermined threshold γ during the period from time t12 to time t14 when the movable part 2a (3) is operating. 20b indicates that the movable part 2a (3) is operating normally. That is, as shown in each of FIGS. 8A and 8B, even when the movable units 2a (1), 2a (2), and 2a (3) operate with overlapping operation periods, the detection unit 20b has the frequency distribution TR (t, By comparing the total expected value TE (t, f) with f) based on the least square method, the abnormality of each movable part 2a can be detected.

  For example, in FIG. 8B, when the weighting coefficient β1 during the period from the time t10 to the time t13 when the movable part 2a (1) is operating is less than a predetermined threshold γ, the detection part 20b 2a (1) is determined to be abnormal.

  FIG. 9 shows an example of monitoring processing in the monitoring apparatus 100 shown in FIG. Steps S200, S210, S220, S230, S240, S250, S260, S270, S280, and S290 are executed when a processor mounted on the monitoring apparatus 100 executes a monitoring program. That is, FIG. 9 shows another embodiment of the monitoring program and the monitoring method. In this case, the estimation units 10b (1) -10b (N), the detection unit 20b, and the control unit 30a illustrated in FIG. 6 are realized by executing a monitoring program. Note that the processing illustrated in FIG. 9 may be executed by hardware installed in the monitoring apparatus 100. In this case, the estimation units 10b (1) -10b (N), the detection unit 20b, and the control unit 30a illustrated in FIG. 6 are realized by circuits arranged in the monitoring device 100.

  In step S200, as described with reference to FIG. 6, the control unit 30a controls the movable units 2a (1), 2a (2), and 2a (3) among the movable units 2a (1) -2a (N). Output a control signal.

  Next, in step S210, the control unit 30a operates the motion function A1 (t) based on the response sequentially received from the movable units 2a (1), 2a (2), and 2a (3) with respect to the output control signal. , A2 (t) and A3 (t) are obtained.

  Next, in step S220, as described with reference to FIGS. 6 and 7, the detection unit 20b generates an electrical signal indicating the vibration of the target device MA1 measured by the measurement unit 3 with a predetermined sampling frequency (16 kilohertz or 22 kHz). Sampling at kilohertz etc.) and obtaining as vibration data.

  Next, in step S230, the detection unit 20b performs a frequency analysis such as FFT on the received vibration data for each predetermined sample unit, for example, to thereby calculate the total frequency of the vibrations of the measured target device MA1. Distribution TR (t, f) is calculated.

  Next, in step S240, the estimation units 10b (1), 10b (2), and 10b (3) transfer the transfer functions H1 (f) of the allocated movable units 2a (1), 2a (2), and 2a (3). ), H2 (f), H3 (f) are read from the storage unit 40, respectively. The estimators 10b (1), 10b (2), and 10b (3) use the operation functions A1 (t), A2 (t), and A3 (t) obtained in step S210, and the read transfer function H1 (f), An overlay operation with H2 (f) and H3 (f) is executed. The estimation units 10b (1), 10b (2), and 10b (3) are expected values E1 (t, f) and E2 (t, f) of the movable units 2a (1), 2a (2), and 2a (3). , E3 (t, f) are estimated.

  Next, in step S250, as described with reference to FIGS. 6 to 8, the detection unit 20b adds the total frequency distribution TR (t, f) and the total expected value TE (t) at a time interval of a predetermined sample unit. , F), the least square method is applied to the amplitude in the frequency axis direction.

  Next, in step S260, as described with reference to FIGS. 6 to 8, the detection unit 20b uses the weighting factors β1, β2, and β3 obtained in predetermined sample units as the movable units 2a (1), 2a ( It is determined whether or not it is equal to or greater than a predetermined threshold γ during the operation period 2) and 2a (3). The detection unit 20b obtains, for each movable unit 2a, the number of samples in a predetermined sample unit from which a weighting coefficient equal to or greater than the predetermined threshold γ is obtained.

  Next, in step S270, as described with reference to FIGS. 6 to 8, the detection unit 20b compares the ratio of the number of samples of the weight coefficient equal to or greater than the predetermined threshold γ with the predetermined ratio in each movable unit 2a. Thus, it is determined whether or not each of the movable parts 2a (1), 2a (2), and 2a (3) is abnormal. For example, if the detection unit 20b determines that an abnormality has occurred in at least one of the movable units 2a (1), 2a (2), and 2a (3) (YES), the process proceeds to step S280. On the other hand, when the detection unit 20b determines that all the movable units 2a (1), 2a (2), and 2a (3) are operating normally (NO), the process proceeds to step S290.

  In step S280, the detection unit 20b outputs an alarm including information on the movable unit 2a in which an abnormality is detected in the output device 4.

  Next, in step S290, the monitoring device 100 determines whether an end instruction has been received via an input device such as a keyboard or a touch panel included in the monitoring device 100, for example. When the monitoring apparatus 100 receives an end instruction (YES), the series of processing ends. On the other hand, if the monitoring apparatus 100 has not received an end instruction (NO), the process proceeds to step S200.

  As described above, in this embodiment, the storage unit 40 uses the transfer function H1 (f) − as information indicating the characteristics of the target device MA1 to transmit the vibration generated by the operations of the movable units 2a (1) -2a (N). HN (t) is stored in advance. The estimation unit 10b (1) -10b (N) is a target device due to vibration generated by the operation of each movable unit 2a from the controlled operation function Aj (t) and transfer function Hj (f) of each movable unit 2a. The expected value Ej (t, f) of the frequency distribution of the vibration of MA1 is estimated. The detection unit 20b is between the measured total frequency distribution TR (t, f) and the expected value TE (t, f) obtained by summing the estimated expected value Ej (t, f) of each movable unit 2a. The weighting coefficient of each movable part 2a obtained by performing the least square method is compared with a predetermined threshold value γ. Thereby, even when the plurality of movable parts 2a operate with overlapping operation periods, the detection unit 20b does not limit the operation of each movable part 2a to a predetermined operation, and detects an abnormality of each movable part 2a. Can be detected.

  The detection unit 20b is controlled based on the least squares process between the measured total frequency distribution TR (t, f) and the estimated total expected value TE (t, f). An abnormality of the M movable parts 2a is detected. For this reason, it is possible to detect an abnormality of the controlled M movable parts 2a by using the measurement part 3 common to the movable parts 2a (1) -2a (N). Thereby, compared with the case where the measurement part 3 is arrange | positioned to each movable part 2a, the number of the measurement parts 3 can be decreased, and cost reduction can be aimed at.

  In addition, the detection unit 20b outputs an alarm including information on the movable unit 2a in which an abnormality is detected in the output device 4. As a result, the administrator of the monitoring device 100 can be encouraged to respond to the movable part 2a in which an abnormality has been detected, and the movable part 2a can be repaired or replaced quickly.

  In FIG. 6, one measuring unit 3 is arranged in the target device MA1, but the present invention is not limited to this. For example, the number of measurement units 3 arranged in the target apparatus MA1 may be two or more and less than N, which is the number of movable units 2a (1) -2a (N). In this case, the measurement part 3 is arrange | positioned in the position which mutually separated in object apparatus MA1. For example, among the plurality of measurement units 3, a measurement unit 3 that can more efficiently measure the vibration generated by the operation of each movable unit 2a than the other measurement units 3 is preset for each movable unit 2a, and the detection unit 20b. Uses vibration data of the target device MA1 measured by the measuring unit 3 set for each movable unit 2a. In this case, when acquiring the transfer function of each movable part 2a, it is preferable that the detection part 20b uses, for example, vibration data of the target device MA1 measured by the measurement part 3 set for each movable part 2a.

  Note that the monitoring process in the monitoring apparatus 100 is not limited to the order of the processes illustrated in FIG. 9. For example, the processes in steps S210 and S240 and the processes in steps S220 and S230 may be performed in parallel. .

  The monitoring device 100 operates as a control device that monitors the operation of the target device MA1 and controls the operation of the target device MA1, but is not limited thereto. For example, a processor or the like included in the target device MA1 functions as a control unit, and the processor or the like of the target device MA1 outputs a control signal to each of the movable units 2a (1) -2a (N). You may receive the response from each movable part 2a. Then, the processor or the like of the target device MA1 may output the received response of the movable unit 2a (1) -2a (N) to the control unit 30a. Further, instead of the control unit 30a of the monitoring device 100, the processor or the like of the target device MA1 calculates the operation function A1 (t) -AN (t) based on the response from the movable unit 2a (1) -2a (N). You may obtain | require and output to estimation part 10b (1) -10b (N).

  Further, although the monitoring device 100 is arranged outside the target device MA1, for example, the monitoring device 100 may be arranged in the target device MA1.

  In addition, in order to acquire the transfer function Hj (f) of each movable part 2a, the operation | movement which the control part 30a performs to each movable part 2a is not limited to single operation decided beforehand, For example, a movable part A plurality of patterns of operations determined in advance for each 2a may be used. And the control part 30a is based on the frequency distribution measured by the measurement part 3 when performing the operation | movement of the some pattern predetermined for every movable part 2a, respectively, and transfer function Hj ( f) may be obtained, and the obtained transfer function Hj (f) may be stored in the storage unit 40.

  Note that the detection unit 20b performs a least-squares process on the total frequency distribution TR (t, f) and the total expected value TE (t, f) during all operation periods of each movable unit 2a. In addition, although it is determined whether or not the weighting coefficient of each movable part 2a is equal to or greater than a predetermined threshold γ, the present invention is not limited to this. For example, the detection unit 20b determines whether or not the obtained weighting factor of each movable part 2a is equal to or greater than a predetermined threshold γ during a period in which the operation of each movable part 2a is in a steady state. This determination may be omitted. Thereby, it is possible to avoid the influence of the impact at the start of the operation of each movable part 2a, noise generated at the end of the operation, etc., and the detection part 20b can avoid erroneous detection of abnormality of each movable part 2a. it can. In addition, since the determination period can be shortened by omitting the determination in a period other than the steady state, the monitoring process in the monitoring apparatus 100 can be speeded up.

  Note that the detection unit 20b detects an abnormality in each movable unit 2a when it is determined in each movable unit 2a that the ratio of the number of samples from which the weight coefficient equal to or greater than the predetermined threshold γ is obtained is smaller than the predetermined rate. However, the present invention is not limited to this. For example, the detection unit 20b determines that there is a possibility of abnormality when there is at least one sample in which the weighting coefficient is determined to be smaller than the predetermined threshold γ in each movable part 2a, and there is a possibility of abnormality. A determination result indicating this may be output to the output device 4. In this case, after the monitoring process illustrated in FIG. 9 is finished, for example, the control unit 30a moves the movable unit 2a within a range in which the movable unit 2a can operate with respect to the movable unit 2a determined to be abnormal. A control signal for inspection or the like may be output. Then, the monitoring apparatus 100 performs a test with higher accuracy than the monitoring process shown in FIG. 9 based on the frequency distribution measured by the measuring unit 3 when the movable unit 2a is operated according to the control signal for testing. May be.

  FIG. 10 shows another embodiment of the monitoring device. Similar to the monitoring device 100 shown in FIG. 6, the monitoring device 100 shown in FIG. 10 operates as a control device that monitors the operation of the target device MA1 and controls the operation of the target device MA1. Among the elements of the monitoring apparatus 100 shown in FIG. 10, those having the same or similar functions as those of the elements of the monitoring apparatus 100 shown in FIG. 10 will be described with reference to FIGS. 7 and 8. FIG.

  The configuration of the target apparatus MA1 is the same as or similar to that of the target apparatus MA1 shown in FIGS. The monitoring device 100 includes N estimation units 10b (1) -10b (N) corresponding to the movable units 2a (1) -2a (N), a detection unit 20c, a control unit 30a, and a storage unit 40.

  For example, the detection unit 20c obtains the expected value E1 (t, f) -EN (t, f) of the frequency distribution, which is a characteristic of the vibration generated in the estimated target device MA1, as the estimation unit 10b (1) -10b ( N), respectively. The detection unit 20c, for example, the estimated frequency distribution expected value E1 (t, f) −EN (t, f) and the total frequency distribution TR (t, f) of the measured vibration of the target device MA1. Is detected, the abnormality of the movable part 2a (1) -2a (N) is detected. In the following, the operation of the detection unit 20c when there are three movable units 2a controlled by the control unit 30a will be described. However, the operation of the detection unit 20c when a plurality of movable units 2a other than three are controlled. The same applies to, and detailed description thereof is omitted.

  For example, the detection unit 20c may calculate the total frequency distribution TR (t, f) illustrated in FIG. 7E and the expected frequency distributions E1 (t, f) and E2 (t) of the estimated vibration of the target device MA1. , F), E3 (t, f), respectively. For example, the detection unit 20c is configured to measure the amplitude in the frequency axis direction between the total frequency distribution TR (t, f) and the expected value E1 (t, f) for each time interval of a predetermined sample unit such as 1024. A cross-correlation process is performed to obtain a cross-correlation coefficient in the movable part 2a (1). The detection unit 20c detects the maximum value of the cross-correlation coefficient of the movable unit 2a (1) and a predetermined threshold value (for each predetermined sample unit) during a period from time t10 to time t13 when the movable unit 2a (1) is operated. For example, 0.5 etc.). The detection unit 20c calculates the number of samples in a predetermined sample unit in which the maximum value of the cross-correlation coefficient of the movable unit 2a (1) is determined to be equal to or greater than a predetermined threshold.

  In addition, for example, the detection unit 20c performs a cross-correlation process on the amplitude in the frequency axis direction for each predetermined sample unit between the total frequency distribution TR (t, f) and the expected value E2 (t, f). Then, the cross correlation coefficient in the movable part 2a (2) is obtained. The detection unit 20c obtains the maximum value of the cross-correlation coefficient of the movable unit 2a (2) calculated in a predetermined sample unit and the predetermined threshold value during the period from the time t11 to the time t15 when the movable unit 2a (2) is operated. Compare. The detection unit 20c calculates the number of samples in a predetermined sample unit in which the maximum value of the cross-correlation coefficient of the movable unit 2a (2) is determined to be equal to or greater than a predetermined threshold. In addition, for example, the detection unit 20c performs a cross-correlation process for the amplitude in the frequency axis direction for each predetermined sample unit between the total frequency distribution TR (t, f) and the expected value E3 (t, f). Then, the cross-correlation coefficient in the movable part 2a (3) is obtained. The detection unit 20c obtains the maximum value of the cross-correlation coefficient of the movable unit 2a (3) obtained in units of a predetermined sample and a predetermined threshold during the period from time t12 to time t14 when the movable unit 2a (3) is operated. Compare. The detection unit 20c calculates the number of samples in a predetermined sample unit in which the maximum value of the cross-correlation coefficient of the movable unit 2a (3) is determined to be equal to or greater than a predetermined threshold.

  For example, in the detection unit 20c, the ratio of the number of samples obtained in each movable part 2a (1), 2a (2), 2a (3) is the total number of predetermined sample units included in the operation period of each movable part 2a. It is determined whether or not a predetermined ratio (for example, 70%, 80%, etc.) or more. For example, when the ratio of the number of samples in which the maximum value of the cross-correlation coefficient equal to or larger than a predetermined threshold is obtained in the movable part 2a (1) in the movable part 2a (1), the movable part 2a (1) Judge that it is operating normally. On the other hand, when the ratio of the number of samples in which the maximum value of the cross-correlation coefficient equal to or greater than the predetermined threshold is obtained in the movable part 2a (1) is smaller than the predetermined ratio, the detection unit 20c causes the movable part 2a (1) to It is determined that an abnormality has occurred. In addition, when the ratio of the number of samples in which the maximum value of the cross-correlation coefficient equal to or greater than a predetermined threshold is obtained in the movable part 2a (2), the detection unit 20c determines that the movable part 2a (2) Judge that it is operating normally. On the other hand, when the ratio of the number of samples from which the maximum value of the cross-correlation coefficient equal to or greater than a predetermined threshold is obtained is smaller than the predetermined ratio, the detection unit 20c has an abnormality in the movable unit 2a (2). judge. In addition, when the ratio of the number of samples in which the maximum value of the cross-correlation coefficient equal to or greater than a predetermined threshold is obtained in the movable part 2a (3), the detection unit 20c determines that the movable part 2a (3) Judge that it is operating normally. On the other hand, when the ratio of the number of samples from which the maximum value of the cross-correlation coefficient equal to or greater than a predetermined threshold is obtained is smaller than the predetermined ratio, the determination unit 20c has an abnormality in the movable unit 2a (3). judge.

  FIG. 11 shows an example of monitoring processing in the monitoring apparatus 100 shown in FIG. Of the processes shown in FIG. 11, processes that are the same as or similar to the processes shown in FIG. 9 are denoted by the same reference numerals, and detailed description thereof is omitted. The processing shown in FIG. 11 is executed by a processor installed in the monitoring apparatus 100 executing a monitoring program. That is, FIG. 11 shows another embodiment of the monitoring program and the monitoring method. In this case, the estimation units 10b (1) -10b (N), the detection unit 20c, and the control unit 30a illustrated in FIG. 10 are realized by executing a monitoring program. Note that the processing illustrated in FIG. 11 may be executed by hardware installed in the monitoring apparatus 100. In this case, the estimation units 10b (1) -10b (N), the detection unit 20c, and the control unit 30a illustrated in FIG. 10 are realized by circuits arranged in the monitoring device 100.

  In the monitoring process shown in FIG. 11, steps S250a, S260a, and S270a are executed instead of steps S250, S260, and S270 of the monitoring process shown in FIG. Steps S200 to S240 are the same as or similar to FIG.

  In step S250a, as described with reference to FIG. 7 and FIG. 10, the detection unit 20c sums up the frequency distribution TR (t, f) and the expected values E1 (t, f), E2 (for each predetermined sample unit). Cross correlation processing is performed on the amplitude in the frequency axis direction between t, f) and E3 (t, f).

  Next, in step S260a, as described with reference to FIG. 10, the detection unit 20c determines that the maximum value of the calculated cross-correlation coefficients of the movable units 2a (1), 2a (2), and 2a (3) It is determined whether or not the movable portion 2a (1), 2a (2), 2a (3) is equal to or greater than a predetermined threshold during the operation period. The detection unit 20c obtains, for each movable unit 2a, the number of samples in a predetermined sample unit in which the maximum value of the cross-correlation coefficient equal to or greater than a predetermined threshold is obtained.

  Next, in step S270a, as described with reference to FIG. 10, the detection unit 20c compares the number of samples from which the maximum value of the cross-correlation coefficient is obtained in each movable unit 2a with a predetermined ratio, and each movable unit It is determined whether or not an abnormality has occurred in 2a (1), 2a (2), and 2a (3). For example, if the detection unit 20c determines that an abnormality has occurred in any of the movable units 2a (1), 2a (2), and 2a (3) (YES), the process proceeds to step S280. On the other hand, if the detection unit 20c determines that all the movable units 2a (1), 2a (2), and 2a (3) are operating normally (NO), the process proceeds to step S290.

  As described above, in this embodiment, the storage unit 40 uses the transfer function H1 (f) − as information indicating the characteristics of the target device MA1 to transmit the vibration generated by each operation of the movable unit 2a (1) -2a (N). HN (t) is stored in advance. The estimation unit 10b (1) -10b (N) is a target device due to vibration generated by the operation of each movable unit 2a from the controlled operation function Aj (t) and transfer function Hj (f) of each movable unit 2a. The expected value Ej (t, f) of the frequency distribution of the vibration of MA1 is estimated. Each of the detection units 20c is obtained by performing a cross-correlation process between the measured total frequency distribution TR (t, f) and the estimated expected value Ej (t, f) of each movable unit 2a. The maximum value of the cross-correlation coefficient of the movable part 2a is compared with a predetermined threshold value. Thereby, even when the plurality of movable parts 2a operate with overlapping operation periods, the detection unit 20c does not limit the operation of each movable part 2a to a predetermined operation, and detects an abnormality of each movable part 2a. Can be detected.

  The detection unit 20c is controlled based on a cross-correlation process between the measured total frequency distribution TR (t, f) and the estimated expected value Ej (t, f) of each movable unit 2a. The abnormality of the M movable parts 2a is detected. For this reason, it is possible to detect an abnormality of the controlled M movable parts 2a by using the measurement part 3 common to the movable parts 2a (1) -2a (N). Thereby, compared with the case where the measurement part 3 is arrange | positioned to each movable part 2a, the number of the measurement parts 3 can be decreased, and cost reduction can be aimed at.

  The detection unit 20 c outputs an alarm including information on the movable unit 2 a in which an abnormality is detected in the output device 4. As a result, the administrator of the monitoring device 100 can be encouraged to respond to the movable part 2a in which an abnormality has been detected, and the movable part 2a can be repaired or replaced quickly.

  Although one measuring unit 3 is arranged in the target device MA1, the present invention is not limited to this. For example, the number of measurement units 3 arranged in the target device MA1 may be two or more and less than N of the movable units 2a (1) -2a (N). In this case, the measurement part 3 is arrange | positioned in the position which mutually separated in object apparatus MA1. For example, among the plurality of measuring units 3, the measuring unit 3 that efficiently measures the vibration generated by the operation of each movable unit 2a is preset for each movable unit 2a, and the detection unit 20c is set for each movable unit 2a. The vibration data of the target device MA1 measured by the measured measuring unit 3 is used. In this case, when acquiring the transfer function of each movable part 2a, it is preferable that the detection part 20c uses, for example, vibration data of the target device MA1 measured by the measurement part 3 set for each movable part 2a.

  Note that the monitoring process in the monitoring apparatus 100 is not limited to the order of the processes shown in FIG. 11. For example, the processes in steps S210 and S240 and the processes in steps S220 and S230 may be executed in parallel. .

  The monitoring device 100 operates as a control device that monitors the operation of the target device MA1 and controls the operation of the target device MA1, but is not limited thereto. For example, a processor or the like included in the target device MA1 functions as a control unit, and the processor or the like of the target device MA1 outputs a control signal to each of the movable units 2a (1) -2a (N). You may receive the response from each movable part 2a. Then, the processor or the like of the target device MA1 may output the received response of the movable unit 2a (1) -2a (N) to the control unit 30a. Further, instead of the control unit 30a of the monitoring device 100, the processor or the like of the target device MA1 calculates the operation function A1 (t) -AN (t) based on the response from the movable unit 2a (1) -2a (N). You may obtain | require and output to estimation part 10b (1) -10b (N).

  Note that the monitoring device 100 is disposed outside the target device MA1, but the monitoring device 100 may be disposed within the target device MA1, for example.

  Note that the operation that the control unit 30a causes each movable unit 2a to perform in order to acquire the transfer function Hj (f) of each movable unit 2a is not limited to a predetermined single operation. A plurality of patterns of operations determined in advance for each 2a may be used. And the control part 30a is based on the frequency distribution measured by the measurement part 3 when performing the operation | movement of the some pattern predetermined for every movable part 2a, respectively, and transfer function Hj ( f) may be obtained, and the obtained transfer function Hj (f) may be stored in the storage unit 40.

  Note that the detection unit 20c performs cross-correlation processing between the total frequency distribution TR (t, f) and the estimated values Ej (t, f) estimated during the entire operation period of each movable unit 2. Further, although it is determined whether or not the maximum value of the cross-correlation coefficient of each movable part 2 is equal to or greater than a predetermined threshold value, the present invention is not limited to this. For example, the detection unit 20c determines whether or not the maximum value of the calculated cross-correlation coefficient of each movable part 2a is equal to or greater than a predetermined threshold during a period in which each movable part 2a is in a steady state. The determination by the period may be omitted. Thereby, it is possible to avoid the influence of the impact at the start of the operation of each movable part 2a, noise generated at the end of the operation, etc., and the detection unit 20c can avoid erroneous detection of abnormality of each movable part 2a. it can. In addition, by shortening the determination period, it is possible to speed up the monitoring process in the monitoring apparatus 100.

  In addition, the detection unit 20c determines that the ratio of the number of samples from which the maximum value of the cross-correlation coefficient equal to or greater than a predetermined threshold is smaller than the predetermined ratio in each movable unit 2a. However, the present invention is not limited to this. For example, the detection unit 20c determines that there is a possibility of abnormality when there is at least one sample in which the maximum value of the cross-correlation coefficient is determined to be smaller than a predetermined threshold in each movable unit 2a. A determination result indicating the possibility may be output to the output device 4. In this case, after the monitoring process shown in FIG. 11 ends, for example, the control unit 30a moves the movable unit 2a within a range in which the movable unit 2a can operate with respect to the movable unit 2a determined to have an abnormality. A control signal for inspection or the like may be output. Then, the monitoring device 100 performs a test with higher accuracy than the monitoring process shown in FIG. 11 based on the frequency distribution measured by the measuring unit 3 when the movable unit 2a is operated according to the control signal for testing. May be.

  FIG. 12 shows an example of the hardware configuration of the monitoring apparatus 100 shown in FIGS. 1, 3, 6 and 10. Note that among the elements shown in FIG. 12, elements equivalent to those shown in FIGS. 1, 3, 6, and 10 are denoted by the same reference numerals, and detailed description thereof is omitted.

  The computer device 200 includes a processor 210, a memory 220, a hard disk device 230, and a robot interface 240. The computer device 200 further includes an output device 4, an input device 250, a network interface 260 and an optical drive device 270. The processor 210, the memory 220, the hard disk device 230, the robot interface 240, the output device 4, the input device 250, the network interface 260, and the optical drive device 270 are connected to each other via a bus. Further, the processor 210, the memory 220, the hard disk device 230, and the robot interface 240 are included in the monitoring device 100.

  The computer apparatus 200 is connected to a bear-type robot CR as the target apparatuses MA and MA1 through the robot interface 240, for example. The processor 210 exchanges information between the sensor 3a included in the robot CR and the movable parts 2a (1) and 2b (2) via the robot interface 240. For example, the processor 210 receives an electrical signal indicating the vibration of the robot CR measured by the sensor 3a such as a vibration sensor or an acceleration sensor via the robot interface 240. Further, the processor 210 outputs an operation control signal to the movable parts 2 a (1) and 2 a (2) such as a servo motor for moving the neck, arms, legs, etc. via the robot interface 240. A response to the control signal is received from the movable parts 2a (1) and 2a (2). In FIG. 12, the sensor 3a is disposed on the head of the robot CR, but the present invention is not limited to this. For example, the sensor 3a may be disposed on the head and the trunk of the robot CR. Moreover, although two movable parts 2a (1) and 2a (2) are arrange | positioned, it is not limited to this, For example, the 1 or 3 or more several movable part 2a may be arrange | positioned.

  The output device 4 includes one or more of a display such as an organic EL or liquid crystal, a speaker, or a light. When the output device 4 detects an abnormality of the movable part 2a (1) or the movable part 2a (2) by the monitoring process by the processor 210, an abnormality has occurred in the movable part 2a (1) or the movable part 2a (2). A warning to that effect is output.

  The input device 250 is, for example, a keyboard, a touch panel, a mouse, or the like. The administrator of the monitoring device 100 operates the input device 250 and, for example, instructs execution or termination of the monitoring process on the movable parts 2a (1) and 2a (2) of the robot CR, or the movable parts 2a (1) and 2a. Input the threshold setting according to (2). Note that the administrator of the monitoring device 100 may reside in a place away from the place where the monitoring device 100 is installed. An administrator who is away from the monitoring apparatus 100 may input an instruction to the monitoring apparatus 100 via the network NW and the network interface 260, for example.

  The optical drive device 270 can be loaded with a removable disk 275 such as an optical disk, and reads and records information recorded on the mounted removable disk 275.

  In addition to the operating system of the computer device 200, the memory 220 stores an application program for the processor 210 to execute monitoring processing. The memory 220 may store in advance the transfer functions, delay times, and the like of the movable parts 2a (1) and 2a (2).

  The application program for executing the monitoring process can be recorded and distributed on the removable disk 275, for example. Then, the application program for executing the monitoring process may be stored in the memory 220 or the hard disk device 230 by attaching the removable disk 275 to the optical drive device 270 and performing the reading process. Further, the computer device 200 may download an application program for executing a monitoring process from the network NW via the network interface 260 and store it in the memory 220 or the hard disk device 230.

  Further, the processor 210 functions as the estimation unit 10 and the detection unit 20 illustrated in FIG. 1 or the estimation unit 10a, the detection unit 20a, and the control unit 30 illustrated in FIG. 3 by executing an application program for monitoring processing. Further, the processor 210 functions as the estimation unit 10b (1) -10b (N), the detection unit 20b, and the control unit 30a illustrated in FIG. 6 by executing an application program for monitoring processing. Alternatively, the processor 210 functions as the estimation unit 10b (1) -10b (N), the detection unit 20c, and the control unit 30a illustrated in FIG. 10 by executing an application program for monitoring processing.

  That is, the monitoring device 100 is realized by the cooperation of the processor 210, the memory 220, the hard disk device 230, and the robot interface 240.

  The application program for the monitoring process is a process for estimating the characteristics of the vibration generated in the robot CR based on the characteristics of the robot CR transmitting the vibration generated by the operation of each movable part 2a and the operation content of each movable part 2a. Includes a program for causing the processor 210 to execute the above. In addition, the application program for the monitoring process is a processor that detects the abnormality of each movable portion 2a by comparing the vibration characteristics of the robot CR measured by the sensor 3a with the estimated vibration characteristics. A program to be executed by 210 is included.

  From the above detailed description, features and advantages of the embodiments will become apparent. This is intended to cover the features and advantages of the embodiments described above without departing from the spirit and scope of the claims. Also, any improvement and modification should be readily conceivable by those having ordinary knowledge in the art. Therefore, there is no intention to limit the scope of the inventive embodiments to those described above, and appropriate modifications and equivalents included in the scope disclosed in the embodiments can be used.

DESCRIPTION OF SYMBOLS 1,30,30a ... Control part; 2, 2a (1) -2a (N) ... Movable part; 3 ... Measurement part; 3a ... Sensor; 4 ... Output device; 10, 10a, 10b (1) -10b (N 20, 20a, 20b, 20c ... detection unit; 40 ... storage unit; 100 ... monitoring device; 200 ... computer device; 210 ... processor; 220 ... memory; 230 ... hard disk device; ... Input device; 260 ... Network interface; 270 ... Optical drive device; 275 ... Removable disk; MA, MA1 ... Target device; CR ... Robot; NW ... Network

Claims (6)

An estimation unit for estimating characteristics of vibration generated in the target device based on characteristics of the target device transmitting vibrations generated by the operation of the movable unit included in the target device and operation details of the movable unit;
A detection unit that detects an abnormality of the movable unit by comparing the estimated vibration characteristic and the measured vibration characteristic of the target device;
A monitoring device comprising: The monitoring device according to claim 1,
The detection unit includes the estimated vibration characteristic and the measured vibration characteristic of the target device in a steady state period in which the operation speed of the movable part is constant among the information indicating the operation content. A monitoring device characterized by comparing. The monitoring device according to claim 1 or 2,
The target device has a plurality of the movable parts,
The detection unit obtains a cross-correlation coefficient of each movable part by performing a cross-correlation process between the measured vibration characteristic of the target device and the vibration characteristic estimated for each movable part. And determining that an abnormality has occurred in the movable part in which the calculated cross-correlation coefficient is smaller than a predetermined threshold value. The monitoring device according to claim 1 or 2,
The target device has a plurality of the movable parts,
The detection unit is configured between a measured vibration characteristic of the target device and a total vibration characteristic obtained by weighting and synthesizing each of the vibration characteristics estimated for each movable part. And calculating a weight corresponding to each of the movable parts by performing a process of least square method, and determining that an abnormality has occurred in the movable part having a value that is less than a predetermined threshold. . Based on the characteristic that the target device transmits the vibration generated by the operation of the movable part included in the target device and the operation content of the movable unit, the characteristics of the vibration generated in the target device are estimated,
By comparing the estimated vibration characteristics with the measured vibration characteristics of the target device, an abnormality of the movable part is detected.
A monitoring program that causes a computer to execute processing. Based on the characteristic that the target device transmits the vibration generated by the operation of the movable part included in the target device and the operation content of the movable unit, the characteristic of the vibration generated in the target device is estimated by the estimation unit,
By detecting the estimated vibration characteristics and the measured vibration characteristics of the target device, the detection unit detects an abnormality of the movable part,
A monitoring method characterized by that.

Images