JPH01295003A - Diagnostic apparatus for rolling machine - Google Patents

Abstract

PURPOSE:To enable abnormity to be always detected by analyzing an input signal and an output signal, judging abnormality, based on the analytical result, and displaying the judgement result, in the apparatus stated in the title for detecting abnormality of a hydraulic pressure depression apparatus for constituting the electric hydraulic pressure servo mechanism of a rolling machine. CONSTITUTION:A wave form processing apparatus 13 for analyzing an input signal and an output signal, a constant memory 15, a data sampling apparatus 16,a first rule memory 17 for judging abnormality from various elements, based on analytical results, judgement calculating apparatus 19, and a display 20 for displaying the judgement results, are provided. Consequently, even if a rolling machine, etc. is operating and complicated signals are contained in the input signal and the output signal, the stable operation and productivity in a rolling machine, etc., can be improved because the abnormity in an electric hydraulic pressure servo system can be detected.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a rolling mill diagnostic device, and more particularly to a diagnostic device suitable for detecting an abnormality in a hydraulic lowering device constituting an electrohydraulic servo mechanism of a rolling mill.

[Conventional technology]

A conventional method for detecting an abnormality in an electro-hydraulic servo system, particularly a method for detecting an abnormality in a hydraulic rolling machine of a rolling mill using an electric servo valve, includes the method described in JP-A-61-88309, which is Each signal of servo current, servo valve opening, and cylinder position is compared with the reference value for the cylinder position command signal, and the polarity of 10 or - in each signal is determined, and a predetermined judgment is made according to the polarity. This detects the location of abnormality based on standards.

[Problem to be solved by the invention]

The conventional method described above is capable of detecting a location where an abnormality occurs when the cylinder position command signal is as simple as shown in equation (1) below.

f (t) = at + b (1) where f (t) cylinder position command signal t: time a, b: constants However, when the rolling mill is in operation, the cylinder position command signal cannot be expressed by a simple function. Yes, the VL becomes a coarse one containing frequency components. Furthermore, signals such as servo current and servo valve opening amount become more complex signals including cylinder position feedback signals, and it is very difficult to compare them with reference signals using a comparison circuit.

SUMMARY OF THE INVENTION An object of the present invention is to provide a rolling mill diagnostic device that can detect abnormalities not only during simple operations such as during maintenance and inspection, but also when the rolling mill is in operation.

[Means to solve the problem]

In order to achieve the above object, the present invention provides a signal capturing means for capturing input signals and output signals of various elements of a device constituting an electro-hydraulic servo system, and an input signal and an output signal captured by the signal capturing means. The apparatus includes an analysis means for analyzing a signal, a determination means for determining an abnormality in the electro-hydraulic servo system based on the analysis result of the analysis means, and a display means for displaying the determination result of the determination means.

Preferably, the analysis means calculates a transfer function from the input signals and output signals of the various elements to obtain data regarding the transfer function, and the determination means has data regarding the transfer function in a normal state set in advance. , the calculated transfer function data is compared with this preset normal transfer function data to determine an abnormality.

The determination means preferably has a preset determination rule, performs a matrix search of the analysis results of the analysis means using the determination rule, and determines an abnormality.

Preferably, the rolling mill diagnostic device of the present invention provides guidance on a failure search method or failure countermeasure based on preset interview rules when the determination means cannot determine an abnormality or when more detailed information is required. The apparatus further includes an inquiry means for displaying information on the display means.

The determination means preferably has a preset transition transfer function for each remaining life, and compares the calculated transfer function with this transition transfer function for each remaining life to determine the remaining life.

The rolling mill diagnostic device of the present invention preferably further includes filter means for removing noise from input signals and output signals of various elements taken into the signal taking means,
This filter means consists of various filters of average method, window method and maximum entropy method.

[Effect]

The input signals and output signals of various elements taken into the signal acquisition means are analyzed by the analysis means, and based on the analysis results, the determination means determines whether there is an abnormality in the electro-hydraulic servo system. Even if the input and output signals include complex signals, it is possible to detect abnormalities in the electro-hydraulic servo system.

When the device constituting the electro-hydraulic servo system is a hydraulic reduction device of a rolling mill, and the element is a servo valve thereof, the analysis means calculates the transfer function of the servo valve from the spool position signal and coil current of the servo valve, The phase or gain gain at one or more frequencies is calculated from the transfer function. The phase or gain gain of the transfer function during normal operation is preset in the determination means, and the abnormality determination is made by comparing the phase or gain gain calculated by the analysis means with the preset phase or gain gain. If a spool stiffness abnormality occurs in the servo valve, a phase delay in the transfer function or a decrease in gain occurs, and the spool stiffness abnormality can be detected from the above comparison. Note that frequencies at one or more delayed phases may be compared.

When the determining means performs a matrix search, an abnormality can be determined by combining the analysis results of a plurality of signals, and the location and contents of the abnormality can be accurately known.

If an inquiry means is provided, even if the judgment means fails to determine an abnormality or more detailed information is required, the operator should carry out the inquiry according to the fault search method or troubleshooting guidance displayed on the display means. By supplementing the information, more accurate and detailed diagnosis can be made.

When the determining means determines the remaining life, it is possible to know the remaining life of the electro-hydraulic servo system element.

When a filter means consisting of various filters is provided, the optimum filter can be selected according to the type of captured signal, and noise can be appropriately removed for each signal, making it possible to perform more accurate analysis.

〔Example〕

Hereinafter, preferred embodiments of the present invention will be described with reference to FIGS. 1 to 7.6 In FIG. , filter 12, waveform processing device 13, A-board 14, constant memory 15 containing sample constant 9 usage data, data sampling device 16, first rule memory 17 containing judgment rules, interview rules? It consists of a second rule memory 18, one judgment calculation device, and a display device 20.

The data acquisition device 10 is composed of a multi-channel AD converter, and includes cylinder position command values, servo currents, etc., which will be described later.
Analog signals such as error voltages are converted into digital signals here and stored in the buffer memory 11. The digital signal passes through a filter 12 that removes disturbance data such as noise, and is taken into a waveform processing device 13, where a transfer function is generated as necessary. In the data sampling device 16, the gain and phase of the transfer function generated by the waveform processing device 13 at the sample constants in the constant memory 15 are calculated. Furthermore, if the input signal and the output signal have a simple relationship, that relationship is calculated. In this way, the waveform processing device 13, constant memory 15, and data sampling device 16 constitute analysis means for analyzing input signals and output signals.

Judgment calculation unit '8. One is the data sampler device 1.
The calculation result in step 6 is compared with the determination data in the constant memory 15, and an abnormal location is determined based on the determination rule in the first rule memory 17. In this way, constant memory 15,
The first rule memory 17 and the judgment calculation device 19 constitute judgment means for determining abnormality of various elements based on the analysis results of the analysis means.The judgment results of the zero judgment calculation device 19 are displayed on the display device 20. be done.

In addition, if an abnormality cannot be determined by one determination calculation device or if more detailed information is required, the display device 20 displays guidance for failure search methods or failure countermeasures based on the inquiry rules in the second rule memory 18. let

Next, the operation between the solid lines will be explained by taking as an example a hydraulic lowering device that constitutes an electro-hydraulic servo mechanism of a rolling mill.

First, the control system of the hydraulic rolling device in the rolling mill will be explained with reference to FIG. In FIG. 2, 21 is a cylinder position command signal for the hydraulic cylinder 22? The cylinder position command signal instructed by the position setting device 21 is added to the cylinder position feedback signal in the adder 23 to generate an error voltage, and the servo amplifier 24 generates a coil proportional to the error voltage. Coil 26 outputs current and creates magnetic flux in coil 26 of servo valve 25
is a spool 28 movable in the axial direction within the sleeve 27;
When a magnetic flux is generated in the coil 26, the coil 26 moves in the axial direction together with the spool 28 due to the magnetic force of the permanent magnet 29 disposed outside the coil 26. When the spool 28 is in the position shown, the pump 30 flows into the hydraulic cylinder 22 through the path indicated by the solid line arrow and pushes up the hydraulic cylinder 22. Further, when the spool 28 moves to the left within the sleeve 27, the oil within the hydraulic cylinder 22 flows into the tank 31 through the path indicated by the dotted line arrow, and the hydraulic cylinder 22 descends. Movement of the spool 28 is detected by a spool position detector 32. Also, the hydraulic cylinder 22
The movement is detected by the hydraulic cylinder position detector 33 and fed back to the cylinder position command signal. By forming a closed loop with the cylinder position command signal, the hydraulic cylinder 22 is operated to always maintain the set position.

The control system of such a hydraulic pressure reduction device is shown in a block diagram as shown in Fig. 3. In Fig. 3, Vref is the cylinder position command signal, Vs is the error signal, iv is the coil current, VO is the spool position signal, and Y indicates the cylinder displacement, and SY indicates the cylinder position feedback signal.

Such failures in the servo system include malfunctions in the arithmetic circuits of the servo amplifier 24 and the adder 23, which prevent the coil current from being generated normally, slowing down the operation of the spool 28, and causing the hydraulic cylinder setting to take too long. There are unfavorable mechanical failures such as a hydraulic shock occurring due to an increase in the inflow speed of hydraulic pressure due to the spool's movement becoming faster.

Furthermore, in the servo amplifier 124, a state in which output is not possible may occur due to the power supply fuse being disconnected. Malfunctions of the servo valve 25 include disconnection of the coil 26, short circuit, and spool 28.
Between the oil and the sleeve 27, there is hardness due to dirt in the oil, and there is wear of the spool 28 and the sleeve 27. In the hydraulic cylinder 22, oil leakage increases due to wear of the paragin between the cylinder 34 and the piston 35, which reduces the efficiency of converting the inflow amount to the stroke amount, and may cause failures that reduce the response speed of the hydraulic cylinder 22. be.

In the rolling mill diagnostic device of this embodiment, the cylinder position command signal VIIEF,
Error voltage VS, coil current iv, spool position signal Vo
- While AD converting and importing the cylinder position feedback signal SY through the data acquisition device 10, the buffer memory 11 is used at every time interval when sufficient data for failure analysis can be obtained.
0 Normally, new data is stored after that time, and the latest data during operation is always stored in the buffer memory 11. Further, in a servo control panel (not shown) of the rolling mill, a failure occurrence signal is output when the cylinder position feedback signal SY is extremely slow with respect to the cylinder position command signal VREF. This failure occurrence signal is captured by the data capture device 10 of this device, and the data capture and update is stopped. At that point, the data before the failure remains in the buffer memory 11. By performing a failure analysis based on this data, the cause of failure occurrence and countermeasures can be taken in a short time.

Specifically, regarding the spool 28, the transfer function of the spool position signal 0 is calculated for the input coil electric K i V. Then, the normal transfer function and the abnormal transfer function are compared in terms of phase delay or gain gain at one or more frequencies. Or compare frequencies at one or more delayed phases. If the difference is larger than a preset value, it is determined that there is an abnormality.

For example, the relationship between the spool position signal Vo and the coil current i can be expressed by the following equation (2).

vo ” K B X G2(S) x i v
(2) Here, KB...servo valve gain constant G2
(S)...The servo valve transfer function G2 (S) depends on the characteristics of the servo valve, and is shown in FIG. 4 when expressed in a Bode diagram with the gain as the vertical axis. In this figure, the solid line shows the normal characteristics.If a spool stiffness abnormality occurs in the servo valve, the gain of the transfer function will decrease, and the Bode diagram will look like the dotted line in Figure 4. In stubbornness, the response in high frequency parts becomes slow. An abnormality can be detected by detecting this change in gain.

Furthermore, if G2(S) is expressed in a Bode diagram with the vertical axis representing the phase, it will be as shown in FIG. Here, if normal, 9
For example, if the 0' phase delay frequency is 150H7, it is 100
Hz, if the limit value of the difference is set to 40 Hz in advance, it can be determined that the spool 28 has become stiff due to dust or the like and is abnormal.

Also, in the power amplifier section, the input to the servo amplifier 24 (
An abnormality can be detected by checking whether the following equation (3) holds true for the error voltage (Vs) and the output (coil current: iV).

iv -K 8 VS
-(3) Here, Vs
......Error voltage iv...Coil current K A...Gain Next, in the adder 23, the cylinder position command signal (
The relationship between VREF) and error voltage (Vs) is as shown in (4) below.
Although the equation is satisfied, anomalies can be detected by examining the equation.

VS = VIIEF -3Y ・(4)
Here, Vs...Error voltage VREF...Cylinder position command signal SY...Cylinder position feedback signal In the hydraulic cylinder 22, the spool position signal is an input, and the cylinder position feedback is an output. Signal SY
Compute the transfer function of If it is normal, the 90' phase lag frequency is, for example, 15 Hz, as shown in Figure 6, but if the piston 34 does not move smoothly due to damage etc., the 90' phase lag frequency may be 5 Hz. For example, the location of the failure can be determined by setting the limit of the difference to 3H2 and determining that there is an abnormality when the limit is exceeded.

Such a servo valve 25, hydraulic cylinder 22, adder 2
3. Determination of abnormality regarding the servo amplifier 24 is performed according to determination rules stored in the first rule memory 17 in advance.

Preferably, the first rule memory 17 is set with an abnormality item determination matrix table as shown in FIG. Determine the content of the abnormality. That is, in the table of FIG. 7, the symbols 1 to 5 are determination coefficients according to the abnormality determination rule, which are obtained by calculating the measured signals. The symbol 1 to 5 is K 1 to 1 (5
This is the boundary value between normal and abnormal for the determination coefficient of , and is a value determined empirically. For example, when the main system determination coefficient is 1≦1, the main system, that is, the adder 23 is normal, and when the power amplifier system is 2≦2, the power amplifier system, that is, the servo amplifier 24 is normal. On the other hand, K1 >k
When l, K2 > K2, the main system and power amplifier system are respectively abnormal. Also, in these combinations,
When all of K 1 to 5 are normal, the hydraulic pressure lowering device and its control system are normal. Regarding abnormalities, for example, if I(1, 3, 4) are abnormal and 2 is normal, the abnormality is a short circuit in the servo valve coil 26.In this way, the contents of each abnormality are determined by combination.

In this determination matrix, 4 is introduced as the servo valve coil system determination coefficient, so the coil voltage VV on the output side of the coil 26 is further detected, and the coil system determination coefficient is determined from the coil voltage iv and the coil voltage VV. It is necessary to find 4.

In addition, the first rule memory 17 is preset with a transition transfer function for each remaining life, and the remaining life can be detected by comparing the calculated transfer function with this transition transfer function for each remaining life. .

That is, in the servo valve 25, since the spool 28 slides inside the sleeve 27 at high speed and frequently, sliding wear naturally occurs, and if a certain limit is exceeded, oil leaks through the worn part. , the oil inflow and outflow efficiency decreases, the operating speed of the spool 28 decreases, and the performance cannot be satisfied.

Conventionally, the servo valve 25 was disassembled, the spool 28 and sleeve 27 were taken out, and the amount of wear was measured, or the servo valve was set on a special test stand and the pressure-gain characteristics were measured to determine the wear limit. was. In any of these methods, it is necessary to remove the servo valve from the rolling mill, which is impossible while the rolling mill is in operation, and requires a large amount of work time. An abnormality is determined by focusing on the relationship between the pressure of the spool 28 and the amount of movement of the spool 28.

In other words, to keep the pressure inside the hydraulic cylinder 22 constant,
It is necessary to always supply oil from the servo valve 25 to compensate for the slight amount of oil leaking from the packing part in the hydraulic cylinder 22.

The oil leak amount is Q, the pressure inside the hydraulic shifter 22 is Pjl,
If the pressure of the pump 30 is PS, and the flow path coefficient due to the operating amount of the spool 28 is f[Vo], these relationships are expressed by equation (5).
). Vo is spool position 13.

Q= f (Vo)J Ps −Pj −(5)f
(v0) naturally changes as the spool 28 wears out, and becomes a transition transfer function f' (Vo).

Therefore, by measuring Q, Ps, and Pj from equation (3), it is possible to determine whether it is a function of the transitive transfer function f' (Vo). Therefore, the wear rate transfer function f'(Vo) is determined in advance, and the actually measured f'(Vo) is compared to estimate the amount of wear, thereby making it possible to detect the remaining life of the servo valve 25 and any abnormalities.

On the other hand, regarding failure determination results, the failure location, failure type, and countermeasures are displayed on the display device 20, but if failure determination cannot be performed or more detailed information is required,
Judgments of a skilled engineer related to the servo system are made into rules and stored in an inquiry rule 18, and based on the inquiry rule, a failure search method or guidance for failure countermeasures is displayed on a display device 20. The researcher supplements information with this device and in the form of an interview using the kaidans shown on the display device 20.
More accurate and detailed diagnosis can be made.

Note that the servo system of a rolling mill has a response of several meters S.
e It is fast from C to more than 10 m5ec, and the control command is several tens of hours.
It changes on the order of z. This change is due to the effects of thickness unevenness, temperature unevenness, hardness unevenness, etc. of the rolled plate material, but the various signals extracted from the servo control system always contain noise, and the transmission of these signals If the function is calculated by 7N, the true transfer function cannot be obtained due to the influence of noise.

Therefore, in this embodiment, a filter 12 is provided in order to remove random noise components after the signal is taken into the buffer memory 11. This filter consists of multiple types of digital filters based on the filter theory of the average method, window method, and maximum entropy method. With this, depending on the type of captured signal! & Select the appropriate filter to remove noise appropriately for each signal,
More accurate analysis becomes possible.

〔Effect of the invention〕

As explained above, according to the present invention, even if the rolling mill is in operation and the input and output signals include complex signals, it is possible to accurately detect abnormalities in the electrohydraulic servo system. This has a great effect on stable operation and productivity improvement of the hydraulic reduction device of the rolling mill.

Furthermore, when performing a matrix search, it is possible to accurately know the details of the abnormality.

If an interview means is provided, a more accurate and detailed diagnosis can be made by supplementing information in the interview format.

When determining the remaining life, it is possible to know the remaining life of the electro-hydraulic servo system elements, making it easier to take countermeasures.

When a filter means consisting of various filters is provided, noise can be appropriately removed for each signal, and more accurate analysis becomes possible.

[Brief explanation of the drawing]

FIG. 1 is a schematic configuration diagram of a rolling mill diagnostic device according to an embodiment of the present invention, FIG. 2 is a system diagram of the hydraulic rolling device of the rolling mill to be diagnosed and its control system, and FIG. is a block diagram of the same hydraulic pressure reduction device and its control system, FIG. 4 is an example of the Bode diagram of the servo valve section, FIG. 5 is another example of the Bode diagram of the servo valve section, and FIG. FIG. 6 is an example of a Bode diagram of the hydraulic cylinder section, and FIG. 7 is a diagram showing a common item determination matrix for the 7-trix search. Explanation of symbols 10...Data acquisition device 12...Filter 16.
...Data sampling device (analysis means) 17...First rule memory (judgment rule) 18...Second rule memory (questionnaire rule) 19...Judgment calculation device (judgment means) 20... Display device applicant Hitachi, Ltd. Agent Patent attorney Haruhi Gho, 8 r+ Shame, 8) ≧ Figure 3 Figure 4 Frequency characteristics Figure 5 Figure 6

Claims (6)

[Claims] (1) A signal importing means for importing input signals and outputs of various elements of the device constituting the electro-hydraulic servo system, an analysis means for analyzing the input signals and output signals taken in, and an analysis result of the analysis means. A rolling mill diagnostic device comprising: a determining means for determining abnormalities in various elements of the electro-hydraulic servo system based on the above information; and a display means for displaying the determination results of the determining means. (2) The analysis means calculates a transfer function from the input signals and output signals of the various elements to obtain data regarding the transfer function, and the determination means has data regarding the transfer function in a normal state set in advance; 2. The rolling mill diagnostic device according to claim 1, wherein abnormality is determined by comparing data of the calculated transfer function and data of the transfer function in a normal state set in advance. (3) The rolling mill diagnostic device according to claim 1, wherein the determination means has a preset determination rule, and searches the analysis result of the analysis means in a matrix according to the determination rule to determine the abnormality. . (4) If the determination means fails to determine an abnormality or more detailed information is required, an inquiry means for displaying guidance on a failure search method or troubleshooting on the display means based on preset inquiry rules. The rolling mill diagnostic device according to claim 1, further comprising: a rolling mill diagnostic device according to claim 1; (5) The analyzing means calculates a transfer function from the input signals and output signals of the various elements, and the determining means has a transition transfer function for each remaining life set in advance, and the calculated transfer function and the transfer function 2. The rolling mill diagnostic device according to claim 1, wherein the remaining life is detected by comparing the transition transfer function for each remaining life. (6) It further comprises a filter means for removing noise from the input signals and output signals of various elements taken into the signal acquisition means, and this filter means uses multiple methods such as an average method, a window method, and a maximum entropy method. The rolling mill diagnostic device according to claim 1, further comprising a filter of different types.