JP7147984B2 - LOAD DETECTION DEVICE, LOAD DETECTION METHOD, AND PROGRAM - Google Patents

Description

The present invention relates to a load detection device and a load detection method for detecting changes in load applied to a mechanical device, and further to a program for realizing these.

A mechanical device with a structure in which a heavy object rotates, such as an antenna device for airport surveillance and a wind power generator, is subjected to a large load due to the repetition of the rotary motion of the heavy object and further by external forces such as wind. Is receiving. In particular, in an antenna device, the reflector, which is a heavy object, rotates in a constant cycle with its center of gravity deviated from its center of rotation (eccentric state), so a very large load is applied to the bearing.

For this reason, in such machinery, the structural members may be exhausted by the load and may eventually be damaged. It is important to understand and detect changes in load. For example, Patent Literature 1 discloses a system for detecting changes in load applied to a mechanical device.

Specifically, Patent Document 1 discloses a system composed of a first stress sensor attached to a rotating member, a second stress sensor attached to a bearing of the rotating member, and a controller. there is In the system disclosed in Patent Document 1, the controller compares the stress of the rotating member detected by the first stress sensor and the stress of the bearing detected by the second stress sensor, and based on the comparison result to extract the stress associated with rotation. Further, the controller detects changes in the load on the rotating member based on the extracted stress.

JP 2009-85893 A

By the way, the stress sensor used in the system disclosed in Patent Document 1 is also called a strain gauge, directly measures strain, and detects stress from the measured strain. However, the strain gauge can basically measure only uniaxial strain depending on the attachment direction. On the other hand, when the direction of the load changes due to the influence of external forces such as wind from all directions, the direction of the stress also changes. Therefore, the strain gauge may not be able to detect stress depending on the direction in which the load is received. Therefore, it becomes difficult to detect stress without being affected by the direction of the load. In other words, in the system disclosed in Patent Document 1, it is difficult to grasp the load state of such a machine without being affected by the direction of the load, and to detect and diagnose changes in the load.

An example of the object of the present invention is to solve the above problems, and in a mechanical device having movable parts, even in situations where loads are applied from various directions, the load applied to the mechanical device can be controlled without being affected by the direction of the load. To provide a load detection device, a load detection method, and a program capable of detecting a change in .

To achieve the above object, a load detection device according to one aspect of the present invention is a device for detecting changes in load applied to a mechanical device having a movable part,
a displacement calculation unit that calculates a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
a load detection unit that compares the calculated time-series change in the displacement with a time-series change in a reference time obtained in advance, and detects a change in the load applied to the mechanical device based on the comparison result;
characterized by comprising

Further, in order to achieve the above object, a load detection method according to one aspect of the present invention is a method for detecting a change in load applied to a mechanical device having a movable part, comprising:
(a) calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
(b) comparing the calculated time-series change in the displacement with the time-series change in the reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result; ,
characterized by having

Furthermore, in order to achieve the above object, a program in one aspect of the present invention is a program for detecting a change in load applied to a mechanical device having movable parts by a computer,
to the computer;
(a) calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
(b) comparing the calculated time-series change in the displacement with the time-series change in the reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result; ,
The program that causes the to run .

As described above, according to the present invention, in a mechanical device having movable parts, even in situations where loads are applied from various directions, changes in the load applied to the mechanical device can be controlled without being affected by the direction of the load. can be detected.

FIG. 1 is a block diagram showing a schematic configuration of a load detection device according to an embodiment of the invention. FIG. 2 is a block diagram more specifically showing the configuration of the load detection device according to the embodiment of the present invention. FIG. 3 is a diagram illustrating components included in the displacement observed on the imaging surface of the imaging device at a certain point when the measurement target region of the object is photographed. FIG. 4 is a diagram schematically showing a two-dimensional spatial distribution (hereinafter referred to as a displacement distribution) of displacement components (δx _ij , δy _ij ) observed in a specific region on an image of the measurement target region. is. FIG. 5(a) is a diagram showing a time series change in displacement calculated by the displacement calculator in the embodiment of the present invention. FIG. 5(b) is a diagram showing a state in which a time window is set for the time-series changes shown in FIG. 5(a). FIG. 5(c) is a diagram showing time-series changes in the reference time used in the embodiment of the present invention. FIG. 6(a) is a graph showing the relationship between the application position of the eccentric load and the amplitude of the oscillation component, and FIG. 6(b) shows the mass of the weight (the magnitude of the eccentric load) and the amplitude of the oscillation component. is a graph showing the relationship between FIG. 7 is a diagram showing changes in peak positions (phases) of time-series changes in displacement. FIG. 8 is a diagram showing the relationship between the amplitude and phase of the oscillation component. FIG. 9 is a diagram showing variation in time-series variation of displacement when the application position of the eccentric load shifts in the radial direction. FIG. 10 is a diagram showing variations in time-series changes in displacement when the application position of the eccentric load is displaced in the circumferential direction. FIG. 11 is a flow chart showing the operation of the load detection device according to the embodiment of the invention. FIG. 12 is a block diagram showing an example of a computer that implements the load detection device according to the embodiment of the invention.

(Embodiment)
A load detection device, a load detection method, and a program according to embodiments of the present invention will be described below with reference to FIGS. 1 to 12. FIG.

[Device configuration]
First, using FIG. 1, a schematic configuration of a load detection device according to an embodiment of the present invention will be described. FIG. 1 is a block diagram showing a schematic configuration of a load detection device according to an embodiment of the invention.

A load detection device 10 according to the present embodiment shown in FIG. 1 is a device for detecting changes in load applied to a mechanical device having a movable part. As shown in FIG. 1 , the load detection device 10 includes a displacement calculator 11 and a load detector 12 .

The displacement calculator 11 calculates the displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion other than the movable portion of the mechanical device. The load detection unit 12 compares the calculated time-series change in displacement with the time-series change at the reference time obtained in advance, and detects the load applied to the mechanical device based on the comparison result.

As described above, in the load detection device 10 according to the present embodiment, minute three-dimensional displacements of the mechanical device are detected not from the stress sensor but from the time-series images. Therefore, according to the load detection device 10, even in situations where loads are applied from various directions in a mechanical device having movable parts, the state of the load applied to the mechanical device can be grasped without being affected by the direction of the load. and can detect changes in load.

Next, the configuration of the load detection device 10 according to the present embodiment will be described more specifically with reference to FIGS. 2 to 10. FIG. FIG. 2 is a block diagram more specifically showing the configuration of the load detection device according to the embodiment of the present invention.

As shown in FIG. 2, first, in the present embodiment, a mechanical device 30 to be detected has a movable portion. Specifically, in the example of FIG. 2, the mechanical device 30 is a radar antenna device (antenna device), a rotating portion (directional antenna) 31 that rotates at a constant cycle, and a support portion that indicates the rotating portion 31. 32. Hereinafter, the mechanical device 30 is also referred to as the radar antenna device 30, and the rotating section 31 is also referred to as the directional antenna 31.

Further, as shown in FIG. 2 , the imaging device 20 is arranged so as to photograph the support portion 32 from a direction perpendicular to the rotation axis of the directional antenna 31 . The imaging device 20 is, for example, a digital camera, and continuously outputs image data as time-series images at a set frame rate.

Furthermore, as shown in FIG. 2, in the present embodiment, the load detection device 10 includes an abnormality determination section 13 and a rotation detection section 14 in addition to the displacement calculation section 11 and the load detection section 12 described above. there is Details of the abnormality determination unit 13 and the rotation detection unit 14 will be described later.

In the present embodiment, based on the time-series images of the measurement target area of the radar antenna device 30, the displacement calculation unit 11 calculates the displacement in the measurement target area in the direction corresponding to the horizontal direction of the time-series images and the time-series image and the displacement in the direction corresponding to the optical axis direction of the imaging device 20 (the direction corresponding to the normal direction of the time-series images). Also, in the present embodiment, the direction corresponding to the rotation axis of the directional antenna 31 is the vertical direction of the time-series images. In this case, the direction corresponding to the horizontal direction of the time-series images and the direction corresponding to the optical axis direction of the imaging device 20 are perpendicular to the rotation axis of the directional antenna 31 .

Hereinafter, the direction corresponding to the horizontal direction of the time-series images will be referred to as the “X direction”, the displacement in the direction corresponding to the vertical direction of the time-series images will be referred to as the “Y direction”, and the optical axis of the imaging device The direction corresponding to the direction (the direction corresponding to the normal direction of the time-series images) is referred to as "Z direction".

Specifically, first, the displacement calculation unit 11 sets a frame at an arbitrary time among the time-series images output from the imaging device 20 as a reference image, and sets other frames as processed images. Then, the displacement calculator 11 searches for a location (coordinates) in the processed image that is most similar to an arbitrary location (coordinates) in the reference image, and calculates the displacement of the specified location (coordinates).

As a method for identifying similar locations, for example, using luminance values of a certain location (coordinates) and coordinates around it, SAD (Sum of Squared Difference), SSD (Sum of Absolute Difference), NCC (Normalized Cross-Correlation), ZNCC (Zero-means Normalized Cross-Correlation), and other similarity correlation functions are used to search for the position (coordinates) with the highest correlation.

In addition, the most similar portion is identified by using the similarity correlation function between the portion with the highest correlation (coordinates) and the portions at the front, rear, left, and right positions (coordinates), and linear fitting, curve fitting, and parabola fitting. It can also be performed by applying a technique such as In this case, the positions (coordinates) of the similar regions can be calculated with higher accuracy and sub-pixel accuracy.

Further, the displacement calculation unit 11 repeats such calculation processing for each location in the processed image, thereby specifying the displacement at each location in the processed image. Furthermore, the displacement calculator 11 acquires the displacement distribution in the measurement target area of the radar antenna device 30 by specifying the displacement at each location for each processed image.

Next, the displacement calculation unit 11 calculates the displacement ΔX in the X direction, the displacement ΔY in the Y direction, and the displacement ΔZ in the Z direction of the measurement target region from the calculated displacement distribution and the imaging information of the imaging device 20. Calculate Further, hereinafter, the displacement ΔX is also referred to as the movement amount ΔX in the X direction of the measurement target area. The displacement ΔY is also expressed as a movement amount ΔY in the Y direction. The displacement ΔZ is also expressed as the amount of movement ΔZ in the Z direction. The imaging information includes at least the size of one pixel of the solid-state imaging device of the imaging device 20, the focal length of the lens, and the imaging distance from the imaging device 20 to the measurement target area (strictly speaking, the distance from the principal point of the lens to the measurement target area). distance), shooting frame rate, and so on.

Here, what displacement components are included in the displacement calculated by the displacement calculation unit 11 will be described with reference to FIGS. 3 and 4. FIG. FIG. 3 is a diagram illustrating components included in the displacement observed on the imaging surface of the imaging device at a certain point when the measurement target region of the object is photographed. FIG. 3 shows a state in which the measurement target area has moved in each direction by the amount of movement (.DELTA.X, .DELTA.Y, .DELTA.Z) before and after the radar antenna device 30 moves due to vibration.

Consider a coordinate system whose origin is the center of the imaging surface of the imaging device 20, that is, the point corresponding to the imaging center, which is the intersection of the optical axis of the lens and the imaging surface. In this coordinate system, consider the observed displacement (δx _ij , δy _ij ) at point A with coordinates (i, j) on the imaging plane of imaging device 20 . Note that the coordinates (i, j) on the imaging surface of the imaging device 20 may be replaced with the coordinates on the captured image.

In the state of FIG. 3, the area to be measured by the radar antenna device 30 has movement amounts (ΔX, ΔY , ΔZ) are generated. The measurement target area moves parallel to the imaging surface of the imaging device 20 by the amount (ΔX, ΔY) of movement in the horizontal direction and vertical direction (X, Y directions) within the screen. Also, it approaches the imaging device 20 by the amount (ΔZ) of movement in the normal direction (Z direction). Therefore, the imaging distance is shortened by the movement amount ΔZ.

Then, as shown in FIG. 3, in addition to the displacement δx caused by the movement amount ΔX of the measurement target area in the horizontal direction (X direction) with respect to the imaging surface of the imaging device 20, a displacement δzx _ij due to the movement amount ΔZ is generated. . Similarly, on the imaging surface of the image pickup device 20, a displacement δzy _ij due to the movement amount ΔZ is also generated in addition to the displacement δy caused by the movement amount ΔY of the image pickup device 20 in the direction perpendicular to the screen (Y direction).

Further, when deformation or displacement (ΔΔX _ij , ΔΔY _ij ) occurs on the surface of the measurement target area due to the radar antenna device 30 receiving some kind of load, the surface displacement component (δδx _ij , δδy _ij ) are produced.

These displacement components are observed as a superposition. In this case, the displacement (δx _ij , δy _ij ) observed at point A can be expressed by the following Equations 1 and 2, as shown in FIG. 4 which will be described later.

Let L be the imaging distance from the principal point of the lens to the measurement target area, f be the lens focal length of the imaging device 20, and (i, j) be the coordinates from the imaging center. Displacement components (δx, δy) associated with movement (Δx, Δy) and displacement components (δzx _ij , δzy _ij ) associated with movement (Δz) in the Z direction are expressed by Equations 3 and 4 below, respectively.

Assuming that all the measurement target areas are moving in the same three-dimensional manner, the displacement components (δx, δy) associated with the movement (Δx, Δy) in the XY directions shown in Equations 3 and 4 are given by is constant regardless of the coordinates of point A shown in . Also, it can be seen that the displacement components (δzx _ij , δzy _ij ) associated with the movement (Δz) in the normal direction increase as the coordinates of the point A move away from the origin. On the other hand, the surface displacement components (δδx _ij , δδy _ij ) of the measurement target region exhibit continuous/discontinuous displacement distributions depending on the position of defects such as cracks on the surface, regardless of the coordinates of the point A.

FIG. 4 is a diagram schematically showing a two-dimensional spatial distribution (hereinafter referred to as a displacement distribution) of displacement components (δx _ij , δy _ij ) observed in a specific region on an image of the measurement target region. is. As shown in FIG. 4, the displacement component (δx _ij , δy _ij ) of each coordinate of the specific region calculated by the displacement calculator 11 is expressed as a displacement vector. In this case, the displacement vector is composed of the displacement components (δx, δy) associated with the movement (Δx, Δy) in the XY directions observed in a uniform direction and size over the entire screen, and the vector group radially from the imaging center of the screen. As a composite component of displacement components (δzx _ij , δzy _ij ) accompanying Z-direction movement (Δz) observed as and surface displacement components (δδx _ij , δδy _ij ) accompanying deformation and displacement of the surface of the measurement target area can be represented.

Consider a method for calculating displacement vector components (δx, δy) associated with movement (Δx, Δy) in the XY directions. As shown in FIG. 4, the displacement components (δx, δy) associated with the movement (Δx, Δy) in the XY directions are basically components observed in a uniform direction and magnitude over the entire screen, like offsets. is. Therefore, from the displacement distribution calculated by the displacement calculation unit 11, displacement components measured at each coordinate of a specific region centered on the imaging center are treated as displacement vectors to which plus and minus are added depending on the direction of displacement. Add all the displacement vectors at each coordinate of interest and take the average. Thereby, the displacement vector components (δx, δy) associated with the movement (Δx, Δy) in the XY directions can be calculated.

This method will be described in detail. First, as shown in Fig. 4, when the displacement distribution is treated as a displacement vector component, displacement vector components (δx, δy) associated with movement (Δx, Δy) in the planar direction and A displacement vector group in which displacement vector components (δzx _ij , δzy _ij ) and surface displacement components (δδx _ij , δδy _ij ) are synthesized is observed. Here, in a specific region centered on the imaging center, as shown in FIG. 4, the displacement vector components (δzx _ij , δzy _ij ) associated with the movement (Δz) in the normal direction are observed as radial displacement vectors. be.

Here, the surface displacement components (ΔΔx _ij , ΔΔy _ij ) are displacement vector components representing deformation/displacement of the surface of the measurement target region. In general, the deformation of the surface is small when no direct stress is applied to the area to be measured. The surface displacement components (δδx _ij , δδy _ij ) are the displacement vector components (δx, δy) accompanying the movement (Δx, Δy) in the plane direction and the displacement vector components (δzxij , δzyij ), so it can be ignored.

In addition, the displacement vector components (δzx _ij , δzy _ij ) associated with the movement (Δz) in the normal direction are radial displacement vectors proportional to the distance to each coordinate (i, j) of the specific area centered on the imaging center. observed as Therefore, when the displacement vector components of each pixel in the specific area around the imaging center are added, they cancel each other out.

From the above, only the displacement vector components (δx, δy) associated with the movement (Δx, Δy) in the XY directions are obtained by adding up the displacement vector components of each pixel in the specific area around the imaging center and calculating the average. can be calculated.

Next, a method for calculating the displacement vector components (Δzx _ij , Δzy _ij ) associated with the movement (Δz) in the normal direction will be described. Consider a state in which only displacement vector components (δzx _ij , δzy _ij ) are generated due to movement (Δz) in the normal direction. The magnitude R(i,j) of the vector is a value proportional to the distance from the imaging center, as shown in Equation 5 below, if the movement amount Δz of the measurement target region is constant within the measurement target region. Become. Also, if k is a constant of proportionality as shown in Equation 6 below, Equation 5 can also be expressed as Equation 7.

On the other hand, the displacement distribution actually calculated by the displacement calculator 11 is composed of composite vector components (δx _ij , δy _ij ) (FIG. 4: thick solid line arrows), as shown in FIG. As can be seen from FIG. 4, the combined vector components (δx _ij , δy _ij ) are the displacement vector components (δzx _ij , δzy _ij ) accompanying movement (Δz) in the Z direction (FIGS. 3 and 4: solid line arrows). , the displacement vector components (δx, δy) accompanying the movement (Δx, Δy) in the XY directions (FIGS. 3 and 4: thick solid line arrows), and the surface displacement components ( δδx _ij , δδy _ij ) (FIGS. 3 and 4: thin solid line arrows).

Of these combined vector components (δx _ij , δy _ij ), the displacement vector components (δx, δy) associated with the previously calculated XY direction movement (Δx, Δy) are subtracted to obtain the Z direction movement (Δz) It corresponds to a composite vector of displacement vector components (δzx _ij , δzy _ij ) and surface displacement components (δδx _ij , δδy _ij ) associated with . _{Therefore , R mes ( i ,} j), these can be expressed as in Equation 8 below, and this value can be calculated.

As shown above, the surface displacement components (δδx _ij , δδy _ij ) are the displacement vector components (δx, δy) associated with movement in the XY direction (Δx, Δy) and the displacement vector components (δx, δy) associated with movement in the Z direction (Δz). In many cases, it can be regarded as sufficiently small compared to the vector components (δzx _ij , δzy _ij ). Therefore, the displacement vector components (δx, δy) associated with the movement in the in-plane direction (Δx, Δy) and the displacement vector components (δzx _ij , δzy _ij ) and proceed with the discussion. In this case, Equation 8 above can be expressed as Equation 9.

In this case, R _mes (i, j) at coordinates (i, j) can be treated as being substantially equal to the displacement vector components (δzx _ij , δzy _ij ) associated with movement (Δz) in the Z direction. At this time, the displacement vector component when the movement amount ΔZ in the Z direction is given is represented by R(i, j) as shown in Equations (6) to (8).

Therefore, from the displacement components (δx _ij , δy _ij ) at each coordinate calculated by the displacement calculation unit 11 and the displacement vector components (δx, δy) accompanying the in-plane movement (Δx, Δy), Using the magnitude R _mes (i, j) of the displacement vector to be obtained, the magnitude R(i, j) of the displacement vector by the displacement vector components (δzx _ij , δzy _ij ) associated with the movement (Δz) in the Z direction is calculated. It is possible to estimate the ratio of enlargement/reduction. Specifically, the scale factor of R(i,j) can be estimated by finding the proportionality constant k that minimizes the evaluation function E(k) shown in Equation 10 below.

Therefore, in the present embodiment, the displacement calculator 11 applies the method of least squares to Equation 10 to calculate the constant of proportionality k. As the evaluation function E(k), in addition to the sum of squares of the difference between R _mes (i ,j) and R(i ,j) shown in Equation 10, the sum of absolute values, other sums of powers, etc. may be used.

Then, the displacement calculator 11 applies the calculated proportionality constant k as a constant indicating the ratio of enlargement/reduction to Equation 7 to calculate the movement amount ΔZ.

As described above, the displacement calculator 11 can obtain the movement amounts ΔX, ΔY, and ΔZ of the measurement target region in the three directions.

Further, the displacement calculator 11 can also calculate the movement amount of the measurement target area with higher accuracy using the calculated movement amounts ΔX, ΔY, and ΔZ of the measurement target area. Specifically, the calculated movement amount ΔZ is substituted into Equation 4 to calculate the displacement vector components (δzx _ij , δzy _ij ) associated with the movement (Δz) in the normal direction. Furthermore, the displacement vector components (δzx _ij , δzy _ij ) associated with the calculated movement (Δz) in the normal direction are subtracted from the displacement vectors (δx _ij , δy _ij ) calculated as the displacement distribution by the displacement calculator 11. Thus, the displacement vector components (δx', δy') associated with the movement (Δx, Δy) in the XY directions are calculated (see Equations 1 and 2 above).

Here, too, the surface displacement components (δδx _ij , δδy _ij ) are the displacement vector components (δx, δy) associated with the in-plane movement (Δx, Δy) and the displacement vector components (Δz) associated with the normal direction movement (Δz). It is calculated using the condition that it can be regarded as sufficiently small compared to the components (δzx _ij , δzy _ij ).

After that, the displacement calculation unit 11 substitutes the displacement vector components (δx', δy') accompanying the calculated movement amount in the XY directions and the movement amount Δz into Equation 3, thereby calculating the XY direction of the measurement target area. Calculate the amount of movement Δx′ and Δy′ in . The movement amounts Δx' and Δy' in the planar direction of the measurement target region calculated in this way are calculated with higher accuracy than the previously calculated movement amounts ΔX and ΔY.

Furthermore, using the calculated movement amounts of the measurement target area in the three directions, Δx' and Δy', the above equation 10 is applied again to calculate Δz', and the movement amounts of the measurement target area in the three directions are calculated. It is also possible to determine Δx', Δy' and Δz'. These values are calculated with higher accuracy than when calculated as the amounts of movement ΔX, ΔY, ΔZ and Δx', Δy', Δz. The above processing may be repeated a predetermined number of times, or may be repeated until convergence within a certain value range.

Further, the movement amounts ΔX and ΔY in the XY directions of the measurement target region and the movement amount ΔZ in the Z direction of the measurement target region, which are calculated by the displacement calculation unit 11, are calculated each time the time-series images are captured. obtained in The (movement amount) for each direction thus obtained represents the time-series change of the displacement in each direction, and can be treated as vibration information with the time interval of imaging as the sampling interval.

The rotation detector 14 detects rotational motion in the radar antenna device 30 . Specifically, the rotation detection unit 14 detects rotational motion based on the displacement information calculated by the displacement calculation unit 11 or based on a signal from a physical device. The “displacement information” is the displacement value for each direction calculated by the displacement calculator 11 . Further, in the present embodiment, when the directional antenna 31 reaches a specific position during rotation, the rotation detection unit 14 detects that fact, outputs a detection signal, and inputs this to the load detection unit 12 . .

When rotational motion is detected based on information on displacement in each direction, the rotation detection unit 14 first detects, for each direction, the displacement calculated by the displacement calculation unit 11 when the directional antenna 31 reaches a specific position. Record the value of the displacement of . Subsequently, the rotation detection unit 14 monitors each value of the displacement for each direction calculated by the displacement calculation unit 11 when the directional antenna 31 is in a rotating state, and the value matches the recorded value. Determine if it is the closest. Specifically, the rotation detection unit 14 determines whether each value is closest to the recorded value by applying the method of least squares to each value of displacement in each direction.

When the rotation detection unit 14 determines that each value of displacement in each direction is closest to the recorded value, the rotation detection unit 14 outputs a signal to the load detection unit 12 at the timing of the determination. The signal output at this time becomes the detection signal described above.

A "physical device" includes a sensor, a switch, and the like. When rotational motion is detected based on a signal from a physical device, a sensor 40 is attached to the support 32 of the radar antenna device 30, as shown in FIG. 2, for example. When the directional antenna 31 reaches a specific position, the sensor 40 detects that and outputs a detection signal. Upon receiving the signal output from the sensor 40, the rotation detection unit 14 outputs the signal to the load detection unit 12 at the timing of the reception. The signal output at this time also becomes the above-described detection signal.

In this embodiment, the load detection unit 12 first detects the rotation detected by the rotation detection unit 14 with respect to the time-series change in the displacement calculated for each direction (X, Y, Z) by the displacement calculation unit 11. Set a time window based on movement. Subsequently, the load detection unit 12 compares the time-series change in the displacement calculated by the displacement calculation unit 11 with the time-series change in the reference time for each set time window for each direction, and Detect such loads.

Specifically, when the directional antenna 31 is at a specific position and a detection signal is output from the sensor 40, the load detection unit 12 uses this detection signal as a trigger to detect the end point of the time window and the next time. Set the starting point of the window. Thereby, the time window is set according to the rotation period of the directional antenna 31 . Also, the setting of the end point and the start point may be performed each time the detection signal is output, or may be performed each time the detection signal is output a specific number of times.

Further, in the present embodiment, in the radar antenna device 30, the directional antenna 31 rotates with its center of gravity shifted from the rotation axis. In other words, when the directional antenna 31 rotates in an eccentric state, a load acts on the directional antenna 31 according to the position of its center of gravity, its weight, and its rotational speed, causing the directional antenna 31 to oscillate. For this reason, the load detection unit 12 focuses on this oscillation component, compares the time-series change of the displacement with the time-series change at the reference time for each set time window, and detects the difference in amplitude and phase. It is also possible to detect the load applied to the radar antenna device 30 based on the extracted difference.

Here, time-series changes in displacement will be described with reference to FIG. FIG. 5(a) is a diagram showing a time series change in displacement calculated by the displacement calculator in the embodiment of the present invention. FIG. 5(b) is a diagram showing a state in which a time window is set for the time-series changes shown in FIG. 5(a). FIG. 5(c) is a diagram showing time-series changes in the reference time used in the embodiment of the present invention. FIGS. 5A to 5C show time-series changes in displacement in the X direction corresponding to the horizontal direction of the time-series images, with the vertical axis representing displacement and the horizontal axis representing time.

As shown in FIG. 5, the displacement fluctuates according to the rotation period of the directional antenna 31 . The time window is set according to the rotation period of the directional antenna 31, and the length of one time window is the time required for the directional antenna 31 to rotate once. The load detection unit 12 detects a load change by comparing the time-series change in the displacement for one rotation shown in FIG. 3(b) with the time-series change at the reference time shown in FIG. 3(c).

Next, load change detection processing by the load detection unit 12 will be described with reference to FIGS. 6 to 10. FIG. Further, in the following description, instead of the directional antenna rotating in an eccentric state, the application position of the eccentric load (distance from the center of rotation) is changed in three steps (0 mm (no eccentricity), 50 mm, 150 mm). The behavior of a rotating table (pedestal) on which a weight is placed will be described. The turntable, which is the supporting part, was set as the measurement target area, and the results of measuring the slight movement of the turntable in accordance with the rotation of the weight are shown below.

FIG. 6(a) is a graph showing the relationship between the application position of the eccentric load and the amplitude of the oscillation component, and FIG. 6(b) shows the mass of the weight (the magnitude of the eccentric load) and the amplitude of the oscillation component. is a graph showing the relationship between In general, when rotation occurs in an eccentric state, centrifugal force acting in the radial direction causes oscillation. The centrifugal force is calculated by "mass x radius x (angular velocity) ² ". Therefore, as shown in FIG. 6A, the amplitude of the swing component increases as the eccentric load applied position increases. Also, as shown in FIG. 6B, the larger the mass of the weight (magnitude of the eccentric load), the larger the amplitude of the swing component. Further, the swing component changes under the influence of external forces such as wind.

FIG. 7 is a diagram showing changes in peak positions (phases) of time-series changes in displacement. As shown in FIG. 7, when the relationship between the trigger position and the weight changes, the phase of the time series change in displacement obtained by image processing changes. Specifically, the application position of the eccentric load differs by 90 degrees between the upper diagram and the lower diagram in FIG. Therefore, in the lower diagram, the peak position of the amplitude of the swing component with respect to the trigger position is shifted by about 1.2 seconds compared to the upper diagram. That is, in the example of FIG. 7, since one period of rotation is 4.66 seconds, a phase shift of about 90 degrees has occurred. As can be seen from FIG. 7, when the position of the weight changes in the circumferential direction, the phase of the time series change in displacement changes.

FIG. 8 is a diagram showing the relationship between the amplitude and phase of the oscillation component. In FIG. 8, the vertical and horizontal axes represent the amplitude of the swinging component, that is, the magnitude of force acting in the radial direction (centrifugal force + other external force). In addition, in FIG. 8, the ● mark indicates the peak of the amplitude at each application position of the eccentric load. The angle θ formed by the line connecting ● to the center of the circle and the line connecting the trigger position to the center of the circle indicates the phase of the peak, in other words, the deviation of the timing of the force acting in the radial direction. ing. Also, the angle θ is proportional to the time it takes for the amplitude to peak after the directional antenna 31 is at the trigger position. From FIG. 8, it can be seen that the occurrence position of the amplitude peak changes when the application position of the eccentric load changes or the magnitude of the external force changes.

In the radar antenna device 30, the rotation center of the directional antenna 31 may be slightly shifted due to abrasion, deformation, etc. of mechanical parts, causing changes in the rotation axis and the position of the center of gravity. In this method, when such wear or deformation occurs, the application position of the eccentric load changes in the radial direction and the circumferential direction. Therefore, even in the radar antenna device 30, the changes shown in FIGS. 6 to 8 can be observed, and the state and change of the load can be detected. Also, in this case, the time-series change of the displacement changes as shown in FIGS. 9 and 10. FIG. FIG. 9 is a diagram showing variation in time-series variation of displacement when the application position of the eccentric load shifts in the radial direction. FIG. 10 is a diagram showing variations in time-series changes in displacement when the application position of the eccentric load is displaced in the circumferential direction.

Therefore, the load detection unit 12 compares the time-series change of the displacement with the time-series change at the reference time, and if the result shown in FIG. An eccentric load is detected. In addition, the load detection unit 12 compares the time-series change of the displacement with the time-series change at the reference time, and if the result shown in FIG. external force).

The abnormality determination unit 13 determines whether an abnormality has occurred in the radar antenna device 30 based on the detected load. Specifically, the abnormality determination unit 13, for example, when comparing the time-series change of the displacement and the time-series change at the reference time, either or both of the difference in the peak of the vibration and the difference in the phase difference is a threshold value or more. If so, it is determined that the radar antenna device 30 is abnormal.

[Device operation]
Next, the operation of load detection device 10 according to the present embodiment will be described with reference to FIG. 11 . FIG. 11 is a flow chart showing the operation of the load detection device according to the embodiment of the invention. 1 to 10 will be appropriately referred to in the following description. Moreover, in this embodiment, the load detection method is carried out by operating the load detection device. Therefore, the description of the load detection method in the present embodiment is replaced with the description of the operation of the load detection device 10 below.

As shown in FIG. 11, first, the displacement calculator 11 acquires image data of time-series images of the measurement target area of the radar antenna device 30 output from the imaging device 20 (step A1). . Next, the displacement calculator 11 uses the image data of the time-series images to calculate the displacement distribution in the measurement target area in the image for each of the X, Y, and Z directions (step A2).

Next, the displacement calculation unit 11 calculates the amount of movement (ΔX, ΔY, ΔZ ) is calculated (step A3). This provides time-series changes in displacement in each direction.

Next, the load detection unit 12 sets a time window with respect to the time-series change in displacement in each direction obtained in step A3, using the detection signal output from the rotation detection unit 14 as a trigger (step A4 ). Next, the load detector 12 compares each waveform separated by the set time window with the time-series change at the reference time to detect the load change in the radar antenna device 30 (step A5).

Next, the abnormality determination unit 13 determines whether or not an abnormality has occurred in the radar antenna device 30 based on the load change detected in step A5, and outputs the determination result (step A6).

By executing step A6, the processing in the load detection device 10 is temporarily terminated, but then step A1 is executed again. That is, in this embodiment, steps A1 to A6 are repeatedly executed.

[Effects of Embodiment]
As described above, according to the present embodiment, the slight swing component of the mechanical device 30 calculated from the time-series images is detected, and the load on the mechanical device 30 is calculated based on the time-series change of the swing component. A change is detected. Therefore, even if the stress fluctuates in a relatively long cycle, changes in the load on the mechanical device 30 can be detected. Also, when external forces such as wind act on the mechanical device 30 and a structure such as a steel tower on which the mechanical device 30 is installed, the swing component changes. Therefore, even in such a case, it is possible to detect the occurrence of a load and changes in the load due to an external force such as wind, without being affected by the direction of the load, by a similar measurement method. Further, in the present embodiment, the abnormality determination is also performed based on the load change, so the burden on the administrator of the mechanical device 30 is greatly reduced.

[Modification]
Next, modifications of the present embodiment will be described below. In the above-described embodiment, the displacement calculator 11 calculates the displacement in each direction from the displacement distribution, but in this modified example, the displacement in each direction is calculated without calculating the displacement distribution.

Specifically, the displacement calculator 11 first compares the processed image with the reference image, and specifies the position of the region with the highest degree of matching with the reference image for each processed image. Further, the displacement calculator 11 calculates the specified position as a displacement d1x in the X direction and a displacement d1y in the Y direction. In the present modification 1 as well, as a method of searching for a target region with the highest matching degree, the above-described similarity correlation functions such as SAD, SSD, NCC, and ZNCC are used to search for the position (coordinates) with the highest correlation. method. Fitting can also be used as a method of searching for an area with the highest matching degree.

Further, the displacement calculator 11 creates an image group (hereinafter referred to as “reference image group”) by enlarging and reducing the reference image at a predetermined magnification in order to calculate the displacement d1z in the Z direction. At this time, the displacement calculation unit 11 sets the center position of the enlarged image and the reduced image of the reference image based on the previously calculated displacement (d1x, d1y) in the XY directions to create the reference image group.

Subsequently, the displacement calculator 11 matches the enlarged image and the reduced image for each processed image, and specifies the enlarged image or the reduced image with the highest degree of matching. An image with a high degree of matching can be identified using, for example, the similarity correlation functions such as SAD, SSD, NCC, ZNCC, etc. described above. Then, the displacement calculation unit 11 identifies an image with the highest degree of similarity, i.e., an image with a high correlation, from among the images forming the reference image group, and determines the enlargement ratio or reduction ratio (hereinafter referred to as “magnification”) of the identified image. ) is calculated as the amount (d1z) indicating the displacement in the normal direction of the specific region.

Further, after identifying the image with the highest matching degree, the displacement calculating unit 11 selects images with magnifications before and after the identified image from among the reference image group, and calculates the degree of similarity between the identified image and the selected image. Calculate the correlation function. Then, the displacement calculation unit 11 may apply a method such as straight line fitting or curve fitting using the calculated similarity correlation function to calculate a magnification that becomes the amount (d1z) indicating the displacement in the normal direction. can. As a result, the magnification (d1z) can be calculated more accurately as an amount indicating the displacement in the normal direction. In this way, the magnification (d1z) is calculated as an amount indicating the displacement in the XY direction (d1x, d1y) and the displacement in the Z direction for each processed image.

In addition, the displacement calculator 11 can perform the above-described processing multiple times in order to increase the accuracy of displacement. Specifically, the displacement calculation unit 11 selects an image corresponding to the magnification d1z from among the images forming the reference image group in consideration of the influence of the previously calculated magnification d1z, and replaces the selected image with a new one. A reference image. Next, the displacement calculation unit 11 compares the processed image with the new reference image, identifies a similar portion in the processed image that is most similar to the new reference image, obtains the position of the similar portion, and determines the position of the similar portion. Detect displacement (d2x, d2y).

Next, based on the newly detected displacement (d2x, d2y), the displacement calculator 11 sets the center position of enlargement or reduction of each image forming the reference image group, and creates a new reference image group. Then, the displacement calculation unit 11 calculates the degree of similarity between the processed image and each image forming the new reference image group, and specifies the image with the highest degree of similarity among the images forming the new reference image group. . After that, the displacement calculator 11 calculates the magnification of the specified image as an amount (d2z) indicating the displacement in the normal direction of the specified area.

Thus, in the first process, displacement (d1x, d1y) is calculated without considering d1z, which is a magnification indicating the displacement in the Z direction, whereas in the second process, The displacement (d2x, d2y) is calculated with the scaling factor d1z taken into account. Therefore, the displacement (d2x, d2y) calculated in the second process is calculated with higher accuracy than the displacement calculated in the first time. In addition, when similar processing is executed a plurality of times, the accuracy of displacement is further improved.

In the above example, the number of repetitions of the process is two, but the number of repetitions is not particularly limited. The number of repetitions may be a preset number, or may be appropriately set according to the results. Alternatively, the process may be repeated until the calculated displacement value reaches the threshold value.

Further, in the following description, the displacement finally obtained in a certain processed image is represented by the displacement (dnx, dny) and the magnification (dnz) which is the amount indicating the displacement in the normal direction. Similar displacement calculation results for time-series images can be treated as time-varying values, so they are denoted as displacement (dnx(t), dny(t)) and magnification (dnz(t)). .

Further, the displacements (dnx(t), dny(t)) in the XY directions are calculated in units of pixels. Therefore, as shown in Equations 11 and 12 below, the displacement calculator 11 calculates the length (Dx, Dy) per pixel of the imaging device of the imaging device 20 in each of the X and Y directions [mm/pixel] is used to calculate the displacement (Δx, Δy) [mm] in each of the X and Y directions. In addition, the length (Dx, Dy) [mm/pixel] per pixel of the image sensor is the pixel pitch (px, py) [mm] of the image sensor, the focal length f [mm] of the lens, and the focal length f [mm] of the lens. It can be calculated from the following Equations 13 and 14 using the distance L [mm] from the principal point to the measurement target area.

Also, the displacement in the Z direction is calculated as a magnification. Therefore, as shown in Equation 15 below, the displacement calculator 11 uses the distance L [mm] from the principal point of the imaging device to the specific region to calculate the displacement Δz [mm] in the Z direction (normal direction). Calculate

Further, the displacements (Δx, Δy, Δz) of the measurement target region obtained in this manner are obtained for each frame in which the time-series images are captured. Therefore, each displacement obtained for each time-series image represents the vibration component of the measurement target region with the reciprocal of the imaging frame rate as the sampling interval.

[program]
The program in this embodiment may be any program that causes a computer to execute steps A1 to A6 shown in FIG. By installing this program in a computer and executing it, the load detection device 10 and the load detection method according to the present embodiment can be realized. In this case, the processor of the computer functions as a displacement calculator 11, a load detector 12, and an abnormality determiner 13 to perform processing.

Also, the program in this embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as one of the displacement calculator 11, the load detector 12, and the abnormality determiner 13, respectively.

Here, a computer that implements the load detection device 10 by executing the program according to the present embodiment will be described with reference to FIG. 12 . FIG. 12 is a block diagram showing an example of a computer that implements the load detection device according to the embodiment of the invention.

As shown in FIG. 12, a computer 110 includes a CPU (Central Processing Unit) 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader/writer 116, and a communication interface 117. and These units are connected to each other via a bus 121 so as to be able to communicate with each other. Further, the computer 110 may include a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array) in addition to the CPU 111 or instead of the CPU 111 .

The CPU 111 expands the programs (codes) of the present embodiment stored in the storage device 113 into the main memory 112 and executes them in a predetermined order to perform various calculations. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). Also, the program in the present embodiment is provided in a state stored in computer-readable recording medium 120 . It should be noted that the program in this embodiment may be distributed on the Internet connected via communication interface 117 .

Further, as a specific example of the storage device 113, in addition to a hard disk drive, a semiconductor storage device such as a flash memory can be cited. Input interface 114 mediates data transmission between CPU 111 and input devices 118 such as a keyboard and mouse. The display controller 115 is connected to the display device 119 and controls display on the display device 119 .

Data reader/writer 116 mediates data transmission between CPU 111 and recording medium 120 , reads programs from recording medium 120 , and writes processing results in computer 110 to recording medium 120 . Communication interface 117 mediates data transmission between CPU 111 and other computers.

Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash (registered trademark)) and SD (Secure Digital); magnetic recording media such as flexible disks; An optical recording medium such as a ROM (Compact Disk Read Only Memory) can be used.

It should be noted that the load detection device 10 according to the present embodiment can also be realized by using hardware corresponding to each part instead of a computer in which a program is installed. Furthermore, the load detection device 10 may be partially implemented by a program and the rest by hardware.

Some or all of the above-described embodiments can be expressed by (Appendix 1) to (Appendix 21) described below, but are not limited to the following descriptions.

(Appendix 1)
A device for detecting changes in load on a mechanical device having moving parts, comprising:
a displacement calculation unit that calculates a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
a load detection unit that compares the calculated time-series change in the displacement with a time-series change in a reference time obtained in advance, and detects a change in the load applied to the mechanical device based on the comparison result;
A load detection device characterized by comprising:

(Appendix 2)
The load detection device according to Appendix 1,
The mechanical device includes a rotating portion that rotates at a constant cycle and a supporting portion that supports the rotating portion, and the imaging device moves the supporting portion from a direction perpendicular to the rotation axis of the rotating portion. are filming,
A load detection device characterized by:

(Appendix 3)
The load detection device according to Appendix 2,
the displacement calculation unit calculates the displacement in three directions, namely, a direction perpendicular to the rotation axis and two directions perpendicular to the direction perpendicular to the rotation axis and perpendicular to each other;
wherein the load detection unit detects changes in the load applied to the mechanical device in the three directions;
A load detection device characterized by:

(Appendix 4)
The load detection device according to appendix 2 or 3,
The load detection unit sets a time window for each period with respect to the calculated time-series change in the displacement, and for each set time window, the calculated time-series change in the displacement and the reference detecting a change in the load applied to the mechanical device by comparing it with the time-series change of time;
A load detection device characterized by:

(Appendix 5)
4. The load detection device according to any one of Appendix 4,
When the rotating part of the mechanical device rotates in a state where the center of gravity and the rotation axis are deviated,
The load detection unit compares the calculated time-series change of the displacement with the time-series change at the reference time for each of the set time windows, extracts differences in amplitude and phase, and extracts the extracted detecting a change in load on the mechanical device based on the difference;
A load detection device characterized by:

(Appendix 6)
The load detection device according to appendix 4 or 5,
further comprising a rotation detection unit that detects rotational motion in the mechanical device;
The load detection unit sets a time window for each cycle based on the rotational motion detected by the rotation detection unit.
A load detection device characterized by:

(Appendix 7)
The load detection device according to any one of Appendices 1 to 6,
further comprising an abnormality determination unit that determines whether an abnormality has occurred in the mechanical device based on the detected load;
A load detection device characterized by:

(Appendix 8)
A method for detecting changes in load on a mechanical device having moving parts, comprising:
(a) calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
(b) comparing the calculated time-series change in the displacement with the time-series change in the reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result; ,
A load detection method characterized by comprising:

(Appendix 9)
The load detection method according to appendix 8,
The mechanical device includes a rotating portion that rotates at a constant cycle and a supporting portion that supports the rotating portion, and the imaging device moves the supporting portion from a direction perpendicular to the rotation axis of the rotating portion. are filming,
A load detection method characterized by:

(Appendix 10)
The load detection method according to Appendix 9,
In step (a), the displacement is calculated in three directions: a direction perpendicular to the rotation axis and two directions perpendicular to the direction perpendicular to the rotation axis and perpendicular to each other. ,
In step (b), detecting changes in the load applied to the mechanical device in the three directions;
A load detection method characterized by:

(Appendix 11)
The load detection method according to Appendix 9 or 10,
In step (b) above, a time window is set for each cycle with respect to the calculated time-series change in the displacement, and the calculated time-series change in the displacement for each set time window; Detecting a change in the load applied to the mechanical device by comparing it with the time-series change at the reference time;
A load detection method characterized by:

(Appendix 12)
The load detection method according to Appendix 11,
When the rotating part of the mechanical device rotates in a state where the center of gravity and the rotation axis are deviated,
In the step (b), for each of the set time windows, the calculated time-series change in the displacement is compared with the time-series change in the reference time to extract differences in amplitude and phase. detecting a change in the load on the mechanical device based on the difference obtained;
A load detection method characterized by:

(Appendix 13)
13. The load detection method according to Appendix 11 or 12,
(c) further comprising detecting rotational motion in the mechanical device;
In (b), setting a time window for each period based on the rotational motion detected in (c);
A load detection method characterized by:

(Appendix 14)
The load detection method according to any one of Appendices 8 to 13,
(d) further comprising determining whether an abnormality has occurred in the mechanical device based on the sensed load;
A load detection method characterized by:

(Appendix 15)
A program for detecting a change in load applied to a mechanical device having moving parts by a computer,
to the computer;
(a) calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
(b) comparing the calculated time-series change in the displacement with the time-series change in the reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result; ,
The program that causes the to run .

(Appendix 16)
The program according to Appendix 15,
The mechanical device includes a rotating portion that rotates at a constant cycle and a supporting portion that supports the rotating portion, and the imaging device moves the supporting portion from a direction perpendicular to the rotation axis of the rotating portion. are filming,
A program characterized by

(Appendix 17)
The program according to Appendix 16,
In step (a), the displacement is calculated in three directions: a direction perpendicular to the rotation axis and two directions perpendicular to the direction perpendicular to the rotation axis and perpendicular to each other. ,
In step (b), detecting changes in the load applied to the mechanical device in the three directions;
A program characterized by

(Appendix 18)
The program according to Appendix 16 or 17,
In step (b) above, a time window is set for each cycle with respect to the calculated time-series change in the displacement, and the calculated time-series change in the displacement for each set time window; Detecting a change in the load applied to the mechanical device by comparing it with the time-series change at the reference time;
A program characterized by

(Appendix 19)
The program according to Appendix 18,
When the rotating part of the mechanical device rotates in a state where the center of gravity and the rotation axis are deviated,
In the step (b), for each of the set time windows, the calculated time-series change in the displacement is compared with the time-series change in the reference time to extract differences in amplitude and phase. detecting a change in the load on the mechanical device based on the difference obtained;
A program characterized by

(Appendix 20)
The program according to Appendix 18 or 19,
to the computer;
(c) further performing the step of detecting rotational motion in the mechanical device ;
In step (b), setting a time window for each cycle based on the rotational motion detected in step (c);
A program characterized by

(Appendix 21)
The program according to any one of Appendices 15 to 20,
to the computer;
(d) further executing a step of determining whether an abnormality has occurred in the mechanical device based on the sensed load ;
A program characterized by

INDUSTRIAL APPLICABILITY As described above, according to the present invention, it is possible to detect changes in the load applied to a mechanical device having movable parts without being affected by the rotation period. INDUSTRIAL APPLICABILITY The present invention is useful for various mechanical devices such as radar antenna devices and wind power generators.

REFERENCE SIGNS LIST 10 load detection device 11 displacement calculation unit 12 load detection unit 13 abnormality determination unit 14 rotation detection unit 20 imaging device 30 mechanical device (radar antenna device)
31 Directional Antenna 32 Support Part 40 Sensor 110 Computer 111 CPU
112 main memory 113 storage device 114 input interface 115 display controller 116 data reader/writer 117 communication interface 118 input device 119 display device 120 recording medium 121 bus

Claims (7)

A device for detecting changes in load on a mechanical device having moving parts, comprising:
Displacement calculation means for calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
load detection means for comparing the calculated time-series change in the displacement with a time-series change in a reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result;
with
The mechanical device includes a rotating portion that rotates at a constant cycle and a supporting portion that supports the rotating portion, and the imaging device moves the supporting portion from a direction perpendicular to the rotation axis of the rotating portion. I am filming
The load detection means sets a time window for each period with respect to the calculated time-series change of the displacement, and for each set time window, the calculated time-series change of the displacement and the reference detecting a change in the load applied to the mechanical device by comparing it with the time-series change of time;
A load detection device characterized by: The load detection device according to claim 1 ,
the displacement calculating means calculates the displacement in three directions, namely, a direction perpendicular to the rotation axis and two directions perpendicular to the direction perpendicular to the rotation axis and perpendicular to each other;
wherein the load sensing means senses changes in load on the mechanical device in the three directions;
A load detection device characterized by: The load detection device according to claim 1 ,
When the rotating part of the mechanical device rotates in a state where the center of gravity and the rotation axis are deviated,
The load detection means compares the calculated time-series change of the displacement with the time-series change at the reference time for each of the set time windows, extracts differences in amplitude and phase, and extracts the extracted detecting a change in load on the mechanical device based on the difference;
A load detection device characterized by: The load detection device according to claim 1 or 3 ,
further comprising rotation detection means for detecting rotational movement in the mechanical device;
The load detection means sets a time window for each period based on the rotational motion detected by the rotation detection means.
A load detection device characterized by: The load detection device according to any one of claims 1 to 4 ,
Further comprising abnormality determination means for determining whether an abnormality has occurred in the mechanical device based on the detected load,
A load detection device characterized by: A method for detecting changes in load on a mechanical device having moving parts, comprising:
(a) calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
(b) comparing the calculated time-series change in the displacement with the time-series change in the reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result;
The mechanical device includes a rotating portion that rotates at a constant cycle and a supporting portion that supports the rotating portion, and the imaging device moves the supporting portion from a direction perpendicular to the rotation axis of the rotating portion. I am filming
In the above (b), a time window is set for each period with respect to the calculated time-series change in the displacement, and for each set time window, the calculated time-series change in the displacement and the reference detecting a change in the load applied to the mechanical device by comparing it with the time-series change of time;
A load detection method characterized by: A program for detecting a change in load applied to a mechanical device having moving parts by a computer,
to the computer;
(a) calculating a displacement of the mechanical device in a specific direction based on time-series images output from an imaging device that captures a portion of the mechanical device other than the movable portion;
(b) comparing the calculated time-series change in the displacement with the time-series change in the reference time obtained in advance, and detecting a change in the load applied to the mechanical device based on the comparison result; ,
and
The mechanical device includes a rotating portion that rotates at a constant cycle and a supporting portion that supports the rotating portion, and the imaging device moves the supporting portion from a direction perpendicular to the rotation axis of the rotating portion. I am filming
In step (b) above, a time window is set for each cycle with respect to the calculated time-series change in the displacement, and the calculated time-series change in the displacement for each set time window; Detecting a change in the load applied to the mechanical device by comparing it with the time-series change at the reference time;
program.

Images