JPS58100734A - Method and device for detecting fault of rotary equipment - Google Patents

Abstract

PURPOSE:To detect a fault of rotary equipment in its early stage by imposing Fourier analysis on a signal showing the condition of the equipment and comparing the pattern of the obtained power spectrum density with a reference pattern. CONSTITUTION:The output S1 of an instrument 2 fitted to rotaty equipment 1 is fetched by a data storage device 5 through input equipment 4 under a command S2 of a controller. A revolutional number signal S3 in the data is sent to a data storage device 6 for correcting the number of revolutions, and other various data signals S4 are sent to a Fourier analysis device 7 to find power spectrum density. A pattern vector generating device 8 uses the peak width of an input from the input equipment 6 and the current number of revolutions to move the peak of a sample pattern X to a frequency position corresponding to the number of revolutions when a reference pattern mu is taught, and then outputs this as a new sample pattern X; and a comparing device 9 calculates the distance to the reference pattern mu in a reference pattern storage device 10 and then compares it with a threshold value.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an abnormality detection method and an abnormality detection apparatus for detecting abnormalities in rotating equipment such as pumps and electric motors.

[Technical background of the invention]

Nuclear power plants use many pumps, electric motors, and other rotating equipment. If these rotating devices break down, the plant must be shut down even if there is no abnormality in the reactor itself, and this has a significant impact on the plant operating rate of the rotating devices.

Therefore, it is desirable to constantly monitor the status of rotating equipment, detect abnormalities early, and prevent their failures.

[Purpose of the invention]

The present invention has been made based on the above-mentioned circumstances, and an object of the present invention is to provide a method and apparatus for detecting abnormalities in rotating equipment, which can detect abnormalities in rotating equipment at an early stage.

[Summary of the invention]

In the present invention, the power spectral density patterns obtained by performing Fourier analysis on signals obtained from displacement meters, velocity meters, accelerometers, etc. attached to rotating equipment or their drive bearings, etc., are calculated based on their power spectral density patterns under normal conditions. The above objective is achieved by comparing with a reference pattern and determining the presence or absence of an abnormality based on the distance from the reference pattern.

In other words, if the sample pattern is represented by an n-dimensional vector
, element Xj of μ.

is defined using μi and the standard deviation σi of Xj in normal times, and when this D is larger than a predetermined threshold α, it is determined to be abnormal.

The basic principle of the present invention is the above-described determination based on a comparison between D defined by equation (1) and the threshold value α, but due to the characteristics of rotating equipment, erroneous determination may occur if the above-described determination is made. There is. In other words, the power spectrum density of a rotating device has a sharp and large peak at a frequency proportional to its rotation speed, so even if the rotation speed of the rotating device slightly deviates from the rotation speed at the time the reference pattern was learned,
D has a large value and may be considered abnormal even though there is actually no abnormality.

In the present invention, the peak of sample pattern is created, this is used as a new sample pattern X, and D is determined using the above equation (1), and a determination is made to prevent erroneous determination.

[Embodiments of the invention]

In FIG. 1, an output signal S1 of an instrument 2 that collectively indicates a displacement meter, a speedometer, an accelerometer, a rotation speed meter, etc. attached to a rotating device 1 or its drive bearing is generated by a diagnosis execution command S1 from a control device 3. The signal is input to the data storage device 5 via the signal input device 4.

Among the various data stored in the data storage device 5, a signal S3 related to the rotation speed of the rotating equipment is sent to the rotation speed correction data storage device 6, and a signal S3 related to other various data is sent to the rotation speed correction data storage device 6.
4 are sent to the Fourier analyzer 7, where they are subjected to 7-lier analysis and their power spectral densities are determined.

The rotation speed correction data storage device 6 stores the rotation speed at the time of creating the reference pattern, an approximate formula expressing the average size of the peak of each vibration mode as a function of the rotation speed, and the width of the peak (depending on the vibration mode). (approximated to a constant value) etc., and outputs the data at a standard pattern creation rotation speed close to the input rotation speed to the pattern vector creation device together with the input rotation speed as a signal S5. . On the other hand, the Fourier analysis result of the signal S4, that is, the signal S6 regarding the power spectrum density is input to the pattern vector creation device 8, and the pattern vector creation device 8 creates a pattern vector as follows.

First, using the peak and current rotational speeds input from the rotational speed correction data input device, it is checked in what mode of peak the element of the sample pattern X is included. Next, each element is moved to the position of the same number as the peak element of the corresponding order in the reference pattern according to the above-mentioned result. Each peak element is multiplied by the ratio of the peak size at the time of standard pattern creation and the current peak size, which was determined by the approximate formula input from the rotation speed correction data storage device 6, and the result is used to calculate the value of each element after correction. value.

The sample pattern X created by each element after this correction is sent to the comparison device 9. The reference pattern μ is inputted to the comparison device 9 from the reference pattern storage device 10, and the comparison device calculates the distance between the input two according to equation (1), and calculates the distance between them using a predetermined threshold value α. Compare with.

If the comparison result is D>α, signals S7 and S8 are sent to the alarm generating device and the output display device, an alarm is generated, and the output display device 12 displays the occurrence of an abnormality. D〈α
At this time, signal 5788 is not output.

As described above, according to the present invention, abnormalities in rotating equipment can be discovered at an early stage, and by taking measures against these abnormalities, it is possible to prevent the occurrence of failures.

Note that the present invention is not limited to the above-mentioned embodiments. For example, by preparing rotation speed correction data and reference patterns for each rotating device, a plurality of rotating devices can be monitored at once.

〔Effect of the invention〕

According to the present invention, abnormalities in various rotating equipment can be detected early,
Since it is possible to prevent such failures, it is possible to stop operations in plants that use a lot of various rotating equipment, such as nuclear reactors, without causing failures in the equipment that forms the core of the plant. It is possible to significantly improve the operating rate of the plant.

[Brief explanation of drawings]

The figure is a block diagram of one embodiment of the present invention. DESCRIPTION OF SYMBOLS 1... Rotating equipment, 2... Instrument, 3... Control device, 4... Signal input device, 5... Data storage device, 6... Rotation speed correction data storage device,
7...Fourier analysis device, 8° pattern vector creation device, 9...Comparison device, 10...Standard (Turn preservation device application agent, patent attorney, Kikuchi Gobu)

Claims (1)

[Claims] Fourier analysis is performed on various signals indicating the status of rotating equipment to obtain power spectrum densities, and an n-dimensional vector whose elements are logarithmically transformed values is created as a primary sample pattern to determine the normality of rotating equipment. The distance between the standard pattern made of the same n-th vector learned in the state and the secondary sample pattern obtained by the following steps a '' c based on the -th sample pattern is compared with a predetermined threshold. Abnormality detection method for rotating equipment for detecting the presence or absence of an abnormality (a): Creation of a reference pattern closest to the rotational speed of the rotating machine when creating the -order sample pattern; determining the belonging of each element in the primary sample pattern by comparison with the peak width at the given rotation speed; Determining the vibration mode, moving the reference pattern of each element to the position of the peak of the corresponding order.Creating the reference pattern from the rotation speed when creating the C0 reference pattern, and the approximate expression between the average size of each order vibration mode peak. Find the ratio of the current peak size to the current peak size,
A secondary sample pattern whose elements are created by multiplying each peak element of the -order sample pattern by this is created. (2) A means for obtaining a signal indicating the state including the rotation speed of the rotating equipment, a signal input device for inputting the signal into a data storage device according to a command from the control device, and a signal input device for receiving the rotation speed signal of the rotating equipment from the data storage device. A rotation speed correction data storage device that stores the input rotation speed at the time of creating the reference pattern, an approximate expression expressing the average size of the peak of each vibration mode as a function of the rotation speed, and the width of the peak, and the data storage device. A Fourier analyzer performs Fourier analysis on various signals other than the rotation speed from the device and calculates the resulting power spectrum density, and the rotation speed is learned using the rotation speed closest to the input rotation speed output by the rotation speed correction data storage device. a pattern vector creation device that corrects the parfous vector density based on the saved data and input rotation speed regarding the reference pattern and the power spectrum density and creates a sample pattern having the correction value as an element; A reference pattern storage device that stores the obtained reference pattern, and a comparison device that calculates the distance between the sample pattern and the reference pattern, compares it with a predetermined threshold value, and generates an output when the distance is greater than the threshold value. An abnormality detection device for rotating equipment, characterized by comprising: