JPH10332477A - Vibration monitor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent an erroneous operation accompanying the pushing of an operation switch (normal learning is performed) in a vibration monitor, wherein a sensor part and a controller part are made to form a unitary body. SOLUTION: In the vibration monitor 10, a vibration sensor, which detects the vibration of a device 15, and a controller part, which performs the specified processing based on the output of the vibration sensor, are provided. Then, the controller part starts learning when the learning-start instruction is given and determines the judging method for normality/abnormality in the actual operating mode based on the waveform information at the normal time of the device. Then, the instruction for starting the learning is issued by pushing a learning start switch 12 attached to a case as a unitary body by a finger F and the like in a specified timing after the finger is separated. Thus, the effect of the vibration generated by the pushing is eliminated, and the learning can be performed based on the vibration generated in the device.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to various devices such as a motor and a hydraulic cylinder, and analyzes the output of a vibration sensor by a computer or the like to determine whether the vibration of the device is normal or abnormal. The present invention relates to a vibration monitoring device, and more specifically, to a vibration monitoring device in which a sensor portion and a controller portion are integrated.

[0002]

2. Description of the Related Art Vibration of various machines and devices to be monitored is detected by a vibration sensor, and the output of the sensor is analyzed and processed by an information processing device (controller unit) such as a computer. There is a vibration monitoring device that is configured to determine whether the vibration monitoring is performed. An example is described in, for example,
There is one described in JP-A-340799.

The device disclosed in this publication operates the information processing device in a learning mode while the device to be monitored vibrates normally. In this learning mode, the vibration waveform from the vibration sensor is sampled over an appropriate period, the vibration waveform is analyzed according to a plurality of predetermined analysis items, and the analysis data of each item is statistically processed.
An algorithm for determining whether the vibration is normal or abnormal is determined based on the processing result (that is, the characteristic of the waveform of the normal vibration).

In the actual operation mode, the vibration waveform from the vibration sensor is sampled as needed, the vibration waveform is analyzed in accordance with each analysis item in the same manner as in the learning mode, and the analysis data of each item is processed in accordance with the determination algorithm. Then, it is determined whether the vibration of the device to be monitored is normal or abnormal.

[0005] The conventional device described above operates normally and normally based on the normal vibration waveform of the device to be monitored input in the learning mode.
An abnormality determination algorithm is automatically generated. Therefore, when a general user applies this vibration monitoring device to a specific device, it is troublesome to analyze characteristics of the normal vibration and abnormal vibration of the device, and to consider an appropriate determination algorithm based on the analysis result. No preparation is necessary. That is, the user only needs to attach the vibration sensor to the device to be monitored and operate it in the learning mode, and in the actual operation mode thereafter, the normal vibration / abnormal vibration is appropriately determined. In this respect, it is a very convenient and excellent vibration monitoring device.

Conventionally, a controller for performing learning in the above-described learning mode and determining whether the operation is normal or abnormal in the actual operation mode and a sensor for detecting vibration are separately formed and connected by signal lines. Was. Recently, the sensor and controller have been built into the same case,
There are also proposals for miniaturization and easy mounting.

[0007]

In the case where the above-mentioned sensor unit and controller unit which are the object of the present invention are integrated, the operation switches are naturally provided on the surface of the same case. Therefore, for example, when the operation switch is pressed to start the learning mode, the case may be vibrated due to the impact of the pressing. If a vibration is generated, learning is performed based on a vibration waveform obtained by superimposing the vibration of the case accompanying the depression of the operation switch on the vibration generated from the object, and accurate learning cannot be performed.

Further, in order to reduce the size of the apparatus, the space for installing the operation switch is narrower than that of the conventional separation type switch. As a result, the number of operation switches that can be installed is limited, the intervals at which the switches are installed are shortened, and the switches themselves must be reduced. As a result, the learning mode start switch may be erroneously pressed. Then, learning starts due to malfunction,
There is a problem that correct learning knowledge accumulated up to that point is erased.

SUMMARY OF THE INVENTION The present invention has been made in view of the above background, and has as its object to solve the above-mentioned problems, and to provide a vibration monitoring device in which a sensor unit and a controller unit are integrated. An object of the present invention is to prevent a malfunction caused by pressing. In particular, normal learning can be performed, and unintended operation (for example, learning start) is not performed.

[0010]

In order to achieve the above object, a vibration monitoring apparatus according to the present invention performs a predetermined process based on an output of a vibration sensor for detecting vibration of a vibration source. It is assumed that a vibration monitoring device is provided in which a controller unit for performing the operation is integrally provided in a case. The controller section starts a learning mode when an instruction to start learning is given, and in this learning mode, a waveform information at normal time when a determination method for determining whether a vibration source is normal or abnormal is input from a vibration sensor. After the learning mode is completed, the operation in the actual operation mode is performed in which it is determined whether the vibration source is normal or abnormal by analyzing the waveform information input from the vibration sensor based on the determined determination method. Use one that has a function. The instruction to start learning is configured to be issued at a predetermined timing after the learning start switch integrated with the case is once pressed and released.

Normally, in this type of monitoring apparatus, a learning mode is executed and learning is started by pressing a learning start switch. On the other hand, in the present invention, learning is not started only by pressing the learning start switch as described above, and learning is then started on condition that the hand / finger or the like is released from the switch. Therefore, vibration is generated by the urging force applied to the case from the hand or the finger generated when the switch is pressed, which affects the vibration waveform (normal waveform) generated in the normal operation (normal waveform is not output). ), There is no problem in the learning process since learning is not started when the button is pressed. After the hand or finger is released from the switch, the vibration sensor can detect the normal waveform because the influence of the vibration by the urging force is eliminated. Therefore,
Since the learning process in the controller unit can be performed based on the normal waveform, a good learning result can be obtained. as a result,
Even in the subsequent actual operation mode, the normal / abnormal judgment is made based on the correct knowledge (learning result), so that the judgment can be made with high accuracy.

The predetermined timing may be immediately after the release (after an elapsed time of 0 or a very short time has elapsed), but preferably after a lapse of a predetermined time from the release of the learning start switch. Item 2). That is, the vibration generated when the switch is released is extremely small, but not zero, as compared with the vibration generated when the learning start switch is pressed. Therefore, a small amount of vibration is generated, and the vibration may affect a normal waveform.
Therefore, by starting learning after a certain period of time has passed and the influence of the vibration generated at the time of the separation has disappeared (including the degree that there is virtually no problem), the vibration sensor can be operated from a normally operating target. Can be extracted, and more accurate learning can be performed.

On the other hand, as another means for solving the problem, a vibration mode and a controller section are integrally provided in a case, and the controller section learns based on a normal waveform. The basic configuration of having the actual operation mode for performing the normal / abnormal judgment based on the judgment method obtained in the above is the same as the invention of each of the above-mentioned claims, but the instruction to start learning is separately provided from the case. A start command signal generated from the formed remote control switch may be used. As a transmission medium of the start command signal by the remote controller, various types such as wireless and infrared can be used.

In this case, since the case and the learning start switch (remote control switch) are formed separately, naturally the case does not vibrate even if the switch is pressed. Therefore, the learning may be started by pressing the switch, or the learning may be started after the switch is once pressed and released as in the above-described respective inventions.

Prior to executing the actual operation mode, the learning mode may be executed every time. However, in a normal case, if the learning mode is executed once and a determination method for the monitored device is obtained. For example, for a certain period thereafter, actual operation is often performed without learning. That is, the opportunity to operate the learning start switch is smaller than the opportunity to operate other operation switches. Further, in the case of a vibration monitoring device in which the vibration sensor and the controller unit are integrated, the size tends to be reduced, and the installation space for the operation switch is also reduced. Therefore, by separately providing a learning start switch that is less frequently used, other operation switches can be efficiently arranged. Further, there is no possibility that the learning start switch is erroneously depressed during the actual operation and the learning is started due to the malfunction.

[0016]

FIG. 1 shows a preferred embodiment of a vibration monitoring device according to the present invention. As shown in FIG. 1, the vibration monitoring device 10 has a built-in vibration sensor and a controller that performs a predetermined process based on an output from the vibration sensor in a cylindrical case 11. further,
On the front surface, an operation switch 12 for giving various commands to the controller unit and a display unit 13 for displaying an output from the controller unit are provided. This display unit 13
Is for outputting a result of determination of normal / abnormal in the actual operation mode, and outputting information such as learning / ending in the learning mode. In order to reduce the size of the apparatus, a lamp such as an LED may be used instead.

A screw 14 protrudes from the back of the case 11. This screw 14 is attached to the device 1 to be monitored.
5 is attached to a screw hole 16 provided in
The vibration monitoring device 10 is fixed to the device. Further, in this example, the signal line 17 is drawn out from the case 11 to transmit the output result to a remote place (such as a monitoring center) so that the state of the device 15 can be confirmed while staying at the remote place. I have.

FIG. 2 shows an example of the internal structure of the vibration monitoring device 10. As shown in FIG. That is, a vibration sensor 20 that detects vibration from the device 15 to be monitored
And a controller unit 21 that acquires the output of the vibration sensor 20 and performs a predetermined process. In the controller section 21, the sensor output is output from the variable gain amplifier 23 and the filters (LPF 24a, BPF 24b, HPF 24c).
It is configured to be input to a computer 26 via a multiplexer 24 and a multiplexer 25. The computer 26 adjusts the gain of the variable gain amplifier 23 as initial processing in the learning mode. The computer 26 samples the sensor output that has passed through each of the three filters 24 while switching the multiplexer 25 at high speed, performs digital conversion, and performs analysis processing as follows.

The computer 26 is a general one-chip
It consists of a microcomputer and a DSP (Digital Signal Processor). In the computer 26, an input unit 27 such as the operation switch 12 and various operation switches and an output unit such as the display unit 13 are linked to transmit and receive various information and commands. The computer 26 performs main information processing as the vibration monitoring device of the present invention. The information processing mainly consists of the learning mode and the actual operation mode. The outline of the processing contents of both modes is shown in the flowcharts of FIGS.

When an instruction to execute a learning mode is given from the input unit 27 (learning start switch 12) to the computer 26,
The process in the learning mode in FIG. 3 is executed. At this time, the monitoring target device 15 is kept in a normal vibration state, and the vibration sensor 20 detects a normal vibration waveform. Then, the learning information up to that point is first cleared (ST101). Next, the gain of the variable gain amplifier 23 is appropriately adjusted (ST1).
02). That is, the output of the vibration sensor 20 is read via the variable gain amplifier 23, the filter 24, and the multiplexer 25, and the variable gain amplifier 23 is controlled so that its input level (output level of the amplifier 23) falls within a predetermined appropriate range. Adjust the gain of.

In the next step 103, the vibration sensor 20
(A normal vibration waveform of the device 15 to be monitored) is sampled at high speed through the variable gain amplifier 23, the filter 24, and the multiplexer 25, and the sampled vibration waveform is used to obtain characteristics such as a maximum value, a minimum value, and an amplitude. Extract quantity. It should be noted that the feature value extraction is performed a predetermined number of times, and an average value and a standard deviation of each item are obtained.

In the next step 104, a determination method (algorithm) for determining whether the vibration is normal or abnormal based on the extracted characteristic amount is generated. Various processes can be used for the generation process of this determination method. And as an example, there is the following.

That is, as shown in FIG. 5, for example, a vibration level (maximum value) obtained when the device 15 is operating normally is acquired a predetermined number of times, and a normal vibration level is obtained from the obtained vibration level. The normal vibration level is the extracted value when the predetermined number of times is one, but since it is usually performed a plurality of times, an average value of each vibration level obtained by performing the plurality of times may be used. , Can be obtained by various methods such as taking the maximum or minimum value. Then, a value twice as large as the normal vibration level obtained in this manner is set as a threshold for judging normal / abnormal. As a result, in the subsequent actual operation, the vibration level is extracted, and when the vibration level becomes twice or more the normal-time vibration level, it is determined to be abnormal.

As shown in FIG. 6, a plurality of threshold values may be set to determine the degree of abnormality. In this case, the type of alarm output changes depending on the degree of abnormality. As an example, threshold value 1 or more and threshold value 2
In the above, various countermeasures can be taken, such as changing the number / color of lamps to be lit, turning on only the lamp when the threshold is 1 or more, and generating a buzzer or the like when the threshold is 2 or more.

Further, depending on the type of the device 15, even if the vibration is small, it may be abnormal. Therefore, as shown in FIG. 7, a normal range is set on the basis of the normal vibration level, and the boundaries are set as thresholds, and both the lower threshold and the upper threshold are determined to be abnormal. You may do so. The normal range is obtained, for example, by obtaining the variance σ of the level obtained in the normal state, setting the range of 2σ as the normal range, obtaining the maximum and minimum of the vibration level, and multiplying the obtained value by a certain coefficient (for example, minimum value × 0). .5, maximum value x 2).

Once the determination method is determined as described above,
The processing shifts to the actual operation mode shown in FIG. That is, the output of the vibration sensor 20 is sampled at a high speed via the variable gain amplifier 23, the filter 24, and the multiplexer 25, and the analysis data of the monitoring item (vibration level in the case of the above example) is obtained at every predetermined analysis unit time. (ST201). The algorithm for calculating the monitoring item is the same as the feature amount extraction in the learning mode, and acquires a predetermined number of data according to a predetermined sampling time and obtains actual data of the monitoring item based on the acquired data.

Next, whether each of the obtained values is within the normal range,
That is, the threshold value set in the learning mode is compared with the threshold value to determine whether the vibration of the device 15 to be monitored is normal or abnormal (ST202). Then, the determination result is output to the output unit 1.
3 (ST203). This series of operations is repeatedly executed at high speed.

** Learning Start Switch and Learning Mode Execution Timing In the integrated vibration monitoring device described above, a learning start command for executing the learning mode is issued when the learning start switch 12 is operated under the following conditions. Like that. This is one of the points of the present invention.

That is, as shown in FIG. 8A, first, the operator presses the learning start switch 12 with the finger F or the like. This press information is given to the built-in computer 26. In this state, the computer 26 does not execute the learning mode. Next, the learning start switch 12 is released as shown in FIG. Information indicating that the switch 12 has returned to its original state is also provided to the computer 26.

As described above, the computer 26 detects that the learning start switch 12 has been pressed once and then released, and executes the learning mode for the first time to perform the learning process (FIG. 3C). Of course, before the learning operation switch is pressed, the device 15 is normally operated. Then
When the learning start switch 12 is released, the vibration generated by the previous pressing is eliminated, and only the vibration based on the normal operation is generated. Therefore, accurate learning can be performed.

The timing when the finger F or the like is released from the learning start switch 12 (FIG. 2B) and the time when learning is actually started (FIG. 2C) are almost simultaneously (immediately). It may be performed, or may be started after a certain time has elapsed. When the learning start switch 12 returns after a certain period of time, the timer can be started, and the learning can be started after the certain period of time.

FIG. 9 shows another embodiment of the present invention. This embodiment is the same as the above-described embodiment in that the vibration sensor and the controller are built in the case 11. Note that the vibration sensor 20 and the controller unit 21 are basically the same as those in the above-described embodiment (FIG. 2), and a detailed description thereof will be omitted. Here, in this embodiment, a remote controller 30 is provided separately from the case 11, and a learning start switch is provided on the remote controller 30.

Thus, when the learning start switch provided on the remote controller 30 is pressed, a learning start command signal generated based on the switch is given to the computer 16 via the receiving circuit (not shown) in the case 11, and the learning mode is set. Is executed (learning starts).

It should be noted that the specific processing in the learning mode and the actual operation mode is not limited to the above-described processing, and in order to make a more accurate determination, it is necessary to increase the number of feature amounts to be extracted. The feature amount to be used is changed according to the device (manually or automatically), and the judgment of normal / abnormal is not a simple threshold processing, but a comprehensive judgment is made using fuzzy or other various judgment algorithms. Of course you can.

As an effect of the third aspect (the embodiment shown in FIG. 9), an effect of not giving a vibration which accompanies the pressing of the start command switch, which is the main point of the present invention, is obviously exerted. However, as a secondary effect, by touching the vibration monitoring device during the actual operation mode, the learning start switch is accidentally pressed and learning is started, and the learning knowledge accumulated so far is cleared. Sometimes it goes away. However, this is only when the sensor unit and the controller unit are touched on the unit body side where the sensor unit and the controller unit are integrated. Even if the remote control method is used, if the learning start switch on the remote control is pressed by mistake, the learning starts erroneously. Will do. Naturally, in the case of the configuration in which the learning start switch is integrated into the main unit, since the size of each switch is small and the interval between other switches is narrow, the learning switch is erroneously set during the actual operation mode. There is a risk of pressing.

In order to solve such a problem, for example, learning is started on condition that the switch is kept pressed for a certain period of time or more (in the case of the embodiment shown in FIG. The learning may be started by releasing the switch afterward). In addition, the system shifts to the learning mode by a certain process, and the learning is started by further performing some process in the state where the learning mode is shifted. As a condition for shifting to the learning mode, a predetermined switch is set for a predetermined time or more. It may be kept pressed (the condition for starting learning thereafter is optional). Further, as described above, if the mode is shifted to the learning mode by pressing the switch for a certain period of time or more, and then the learning is started by pressing the switch at least once shortly, malfunction can be more reliably prevented. Furthermore, as a condition for starting learning from the learning mode, the learning start switch may be kept pressed for a certain period of time or more (the condition for shifting to the learning mode is arbitrary).

The switch can be a double-action switch. As is well known, this double action switch is lightly pressed,
It is possible to take two states of being pushed down, and it is convenient to be able to execute different operations in each state. In particular, in order to reduce the size of the device, it is preferable that the switch can be shared. If the state where the switch is pushed down to the second stage (below) is used for learning start when such a switch is used, the learning is not started only with the state of the first stage which is touched and lightly pressed by the switch, so that the learning is not started. The likelihood of starting learning is reduced. Then, in this case, if learning is started only when the state of pressing the second stage is pressed for a certain period of time or more, erroneous learning start can be suppressed more reliably.

Further, a plurality of switches may be provided, and learning may be started on condition that at least two predetermined switches are simultaneously pressed. This way,
Even if one switch is pressed by mistake, the condition that two or more switches are pressed at the same time is not satisfied, so that the possibility of starting learning by mistake is reduced as much as possible.

[0039]

As described above, according to the present invention, even in a vibration monitoring apparatus in which a sensor section and a controller section are integrated, learning is started by once pressing and then releasing the learning start switch. Therefore, the influence of the vibration generated when the learning start switch is pressed can be suppressed, and the learning can be performed based on the vibration generated from the normally operating target. When using the remote control switch,
Since the learning start switch is formed separately from the case, even if the learning start switch is pressed, vibration does not occur on the case side.

Therefore, a favorable learning result can be obtained since there is no influence (less) on the vibration generated by operating the switch. Therefore, even in the actual operation mode after that,
Since normality / abnormality can be determined based on correct learning knowledge (determination method), highly accurate determination / monitoring can be performed.

[Brief description of the drawings]

FIG. 1 is a perspective view showing a preferred embodiment of a vibration monitoring device according to the present invention.

FIG. 2 is a block diagram showing an internal configuration of the vibration monitoring device.

FIG. 3 is a flowchart illustrating a learning mode.

FIG. 4 is a flowchart illustrating an actual operation mode.

FIG. 5 is a diagram illustrating an example of a creation process of a determination method.

FIG. 6 is a diagram illustrating an example of a creation process of a determination method.

FIG. 7 is a diagram illustrating an example of a creation process of a determination method.

FIG. 8 is a diagram illustrating a relationship between operation of a learning start switch, which is a main part of the present invention, and learning start.

FIG. 9 is a diagram showing another embodiment of the vibration monitoring device according to the present invention.

[Explanation of symbols]

 Reference Signs List 10 vibration monitoring device 11 case 12 learning start switch 15 device (vibration source) 20 vibration sensor 21 controller unit 30 remote control

Claims (3)

[Claims] A vibration sensor for detecting vibration of a vibration source;
In a vibration monitoring device integrally provided in a case with a controller unit that performs a predetermined process based on the output of the vibration sensor, the controller unit starts a learning mode when a learning start instruction is given. In the learning mode, a determination method for determining whether the vibration source is normal or abnormal is determined based on the normal waveform information input from the vibration sensor, and after the learning mode ends, based on the determined determination method. It has a function of operating in an actual operation mode to determine whether the vibration source is normal or abnormal by analyzing the waveform information input from the vibration sensor, and the instruction to start learning is attached integrally to the case. A vibration monitoring device characterized in that the learning start switch is pressed once, and is emitted at a predetermined timing after being released. 2. The vibration monitoring device according to claim 1, wherein the predetermined timing is after a lapse of a predetermined time from when the learning start switch is released. 3. A vibration sensor for detecting vibration of a vibration source,
In a vibration monitoring device integrally provided in a case with a controller unit that performs a predetermined process based on the output of the vibration sensor, the controller unit starts a learning mode when a learning start instruction is given. In the learning mode, a determination method for determining whether the vibration source is normal or abnormal is determined based on the normal waveform information input from the vibration sensor, and after the learning mode ends, based on the determined determination method. It has a function of operating in an actual operation mode to determine whether the vibration source is normal or abnormal by analyzing the waveform information input from the vibration sensor, and the instruction to start learning is formed separately from the case. A vibration monitoring device, which is a start command signal issued from a remote control switch.