JPH0728502A - Plant controller - Google Patents

Abstract

PURPOSE:To secure the stable control of an object plant by properly discriminating the abnormal state including a symptom in accordance with measured data of a sensor to automatically and quickly switch the sensor to an alternative means. CONSTITUTION:This plant controller is provided with a sensor 9 which is installed in a required place of the object plant and measures the process state of the object plant, a data analysis means 32 which subjects measured data from this sensor 9 to wavelet conversion to take out analysis data for detection of a discontinuous point, an abnormality discriminating means 34 which uses analysis data from the data analysis means 32 to discriminate whether the sensor is normally operated or not, and a control switching means 35 which switches the sensor judged to be abnormal to an alternative measuring means or switches the control system to another control system using no sensors at the time of discriminating the abnormality by the abnormality discriminating means 34.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plant control device used for controlling ferrous / non-ferrous rolling plants, petrochemical plants, papermaking plants and various other plants, and in particular, an abnormality of a sensor for detecting the state of a target plant. Further, the present invention relates to a plant control device that stably controls a target plant with respect to a similar state.

[0002]

2. Description of the Related Art Industrial fields applicable to this type include iron and non-ferrous ironmaking, steelmaking, rolling and other processes, petrochemical-related refining, distillation and reaction processes, papermaking-related digestion, papermaking, There are processes such as recovery, and various sensors are installed in the plant that handles these processes. In addition, various sensors are similarly installed in the control of the power generation / electric power plant.

By the way, the sensors installed in such various plants are very important not only for measuring the state of the plant but also for controlling the target plant by obtaining the manipulated variable from the measured data. Have a role.

Therefore, when the measurement result cannot be used due to the abnormality of the sensor, an alternative measuring means for estimating or predicting the state of the plant from the measurement results of other sensors is used. The control accuracy is worse than that of.

Further, the location where the sensor is installed may be in a severe environment for the sensor to operate normally, and the sensor does not always operate normally. For example, the load cell installed in the looper between the stands in the continuous hot rolling mill for steel is useful as a sensor for measuring the tension of the steel sheet, but since water such as roll cooling water is often used, The sensor often becomes abnormal due to flooding.

[0006]

By the way, when the sensor becomes abnormal as described above, it is obvious that the stability of the plant cannot be maintained if the abnormal measurement data is directly used for controlling the plant.

Usually, noise is often added to the measured value of the sensor, but it is difficult to judge whether or not the noise is abnormal due to the noise, and a sign that the sensor is abnormal until it becomes abnormal, for example, a moment. In many cases, the signal interruption or the intermittent signal etc. are generated, but it is necessary to take an early action before an abnormality occurs.

Therefore, in the case of a sensor that is important for control, an operator or an observer monitors it, and when it is judged to be abnormal, it is manually switched to another sensor, or the sensor which has become abnormal is disconnected, while visually observing. While operating the plant, for example, the tension state of the rolled material, the plant operation is continued by manual control.

Therefore, when the sensor is abnormal, the operator or the supervisor switches to the alternative measuring means or the different control method, so that the control of the plant becomes unstable due to the delay of the switching timing, or the state of the plant cannot be accurately observed. In the case of environment, there is a problem that manual control cannot be applied.

Further, although it is necessary to take measures such as repair and replacement for the abnormality of the sensor, it is in various aspects to always take a monitoring and maintenance system for the sensor which does not know when it will break down. Not desirable.

The present invention has been made in view of the above circumstances, and appropriately determines an abnormal state including a sign from measurement data of a sensor installed in a target plant, and automatically and quickly substitutes for another. An object of the present invention is to provide a plant control device that ensures stable control of a target plant by using control that does not use measuring means or an abnormality sensor.

[0012]

In order to solve the above-mentioned problems, the invention corresponding to the claims is provided with a sensor which is installed at a required location of a target plant and measures the process state of the target plant, and this sensor. Data analysis means for extracting analysis data for finding discontinuity points by wavelet transform of measurement data from, and abnormality determination for determining whether or not the sensor is normally operating using the analysis data by this data analysis means Means and a plant control device provided with a control switching means for switching to a measuring means which is a substitute for the abnormality sensor when it is judged to be abnormal by this abnormality judging means, or for switching to a control system which does not use the abnormality sensor. is there.

In the abnormality determining means, the neural network is provided with the signal pattern after the wave transformation and the output value at various abnormal times of the sensor to give the connection weight coefficient between the layers of the neural network. While inputting the learning means for determining and the analysis data of the data analyzing means at the time of actual operation of the target plant to the neural network, the connection weight coefficient is set in the neural network to determine whether the sensor is abnormal or not. It has a judgment means for judging.

[0014]

Therefore, according to the invention corresponding to the claims, by taking the above-mentioned means, the measurement data from the sensor is used for controlling the target plant and is also introduced into the data analyzing means. Is performed to create analysis data for finding a discontinuity point and send it to the abnormality determining means.

In this abnormality judging means, the coupling weight coefficient is stored in advance by wavelet transforming the signals emitted by the sensor at various abnormal times. Therefore, the pattern of the stored coupling weight coefficient and the actual sensor are used. By comparing the measurement data with the wavelet-transformed signal, it is possible to accurately determine whether or not the sensor is abnormal.

When the abnormality determination means determines that the sensor is abnormal, the control switching means automatically and quickly switches to an alternative sensor or to a control method that does not use the abnormality sensor. It is possible to ensure the stability of plant control.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of a plant control apparatus according to the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram applied to a tension control between two stands of a continuous hot rolling mill, for example, as a first embodiment of a plant control device.
Although the continuous rolling mill is composed of about 4 to 7 stands, the generality is not lost even if only two stands are considered.

In FIG. 1, reference numeral 1 denotes a rolled material which travels in the direction of the arrow shown in the drawing. The rolled material 1 is rolled by an i-stand rolling mill 2 and an i + 1-stand rolling mill 3, where a continuous rolling mill is an n-stand rolling mill. Then, i = 1 to n-1
Becomes These i-stand rolling mill 2 and i + 1-stand rolling mill 3 are driven by a main motor (hereinafter, referred to as i-stand main machine) 4 and a main motor (hereinafter, referred to as i + 1-stand main machine) 5, respectively. Reference numerals 6 and 7 are speed control devices (ASR: Automatic Speed Regulator) in the i-stand main engine 4 and the i + 1-stand main engine 5, respectively.

Reference numeral 8 is a looper, 9 is a load cell or strain gauge (hereinafter referred to as a sensor) installed in the looper 8, 10 is a looper electric motor for driving the looper 8, and 11 is a drive device for the looper electric motor 10. Speed control (A
SR) or current control (ACR: Automatic Current)
Regulator). 12 is a looper angle measuring device for measuring the looper angle, 13 is a sensor 9,
Alternatively, it is a tension calculation means for obtaining the tension of the rolled material 1 from the current of the looper electric motor 10. These looper angle measuring devices 12
The looper angle θ measured in step S1 and the tension t _f obtained by the tension calculator 13 are the tension controller (ATC: Automati).
c Tension Controller 14).

The tension control device 14 uses the looper angle θ and the tension t _f to obtain a speed command value N _REF to control the speed of the i-stand main engine 4 via the speed control device 6, and also to control the speed. Command value N _LREF or current command value I
Loop motor 1 via drive device 11 for _LREF
0 speed control or current control is performed.

Reference numeral 15 denotes a plate thickness gauge installed between stands, which may not be installed depending on the type of rolling mill or stand. When the thickness gauge 15 is installed, the thickness control of the i stand and i + 1 stand (AGC: Auto
The matic gage control devices 16 and 17 are plate thickness gauges 15.
The roll gap control signal is calculated based on the measured thickness of the roll gap (APC: Automatic Posit).
Ion Controller) 18 and 19 to control the plate thickness.

On the other hand, when the plate thickness gauge 15 is not installed, the rolling load is measured by the load cells 20 and 21, and the plate thickness directly below the stand is calculated by the gauge meter type together with the roll gap opening signal, and the calculation is performed. The so-called gauge meter AGC method is used, in which the plate thickness is controlled by using the measured plate thickness value.

Now, the tension control is performed by using the conventional control method of adjusting the speed of the main machine 4 so as to keep the torque received by the looper 8 constant and the tension value t _f obtained from the measured value of the load cell or the like. Interference control, LQ (Linear Quadrati
Control methods such as c) control, ILQ (Inverse LQ) control, and H 2 control are used. In the conventional control method, since the tension value obtained by calculation is not used, a sensor for tension detection is unnecessary, but the actual tension value is unknown, and mutual interference in the looper system remains. There is a disadvantage in that a fast response cannot be expected because the control is performed as it is.

On the other hand, in the control system in which the tension control is performed by using the tension value calculated from the measurement data of the sensor 9, the normal tension detection is a condition for good control, and It is superior to the conventional control system in high-speed response. Therefore, the plant is fast and
For stable control, it is very important to control tension using sensor measurement data.

However, since the sensor 9 for detecting the tension is attached to the looper 8, water for cooling the rolls of the rolling mill, water for cooling the looper 8 or the scale of the rolled material 1 between the stands is used. In many cases, a descaler that injects water at a high pressure to remove the film, water for cooling the rolled material 1, and the like enter the sensor 9. In particular, since the sensor 9 such as a load cell or a strain gauge often converts mechanical stress into an electric signal and outputs the electric signal, it cannot be used in many cases when the water is flooded. Further, the sensor 9 is not so reliable due to the high heat of the rolled material 1, the life of the sensor itself, and the like.

Therefore, it is necessary to constantly monitor the output signal from the sensor 9 and promptly replace the sensor when an abnormality occurs, or take countermeasures when a sign of abnormality appears. On the other hand, even if it is determined that the sensor is abnormal during rolling, it is difficult to compensate the operation of the control device by using the measured value of the sensor, and it is difficult to determine whether or not the abnormality is occurring.

Therefore, in the present apparatus, a sensor switching means 31 is provided between the sensor 9 and the looper electric motor 10 and the tension calculating section 13, and further, data analysis is carried out by wavelet transforming the measurement data by the sensor 9 for data analysis. Means 32, learning support means 33 for creating a signal wavelet-transformed from various abnormal signals of the sensor 9, and giving it as a teacher signal of a neural network for learning, and an analysis result by the data analysis means 32. And an abnormality determination means 34 for determining whether the sensor 9 is operating normally or abnormal using the teacher signal of the learning support means 33, and when the abnormality determination means 34 determines an abnormality , Sensor 9
Is output, and at the same time a switching control signal is sent to the sensor switching means 31 to switch the sensor 9 to the alternative measuring means side.

Next, the operation of the portion of the apparatus constructed as described above, which is particularly related to the abnormality determination of the sensor 9, will be described. First, the measurement data measured by the sensor 9 is introduced into the data analysis means 32. At this time, when the sensor 9 becomes abnormal, the measured data can be generally considered as a discontinuous signal. Therefore, the data analysis means 32 creates analysis data such that a discontinuous point can be clearly obtained by performing a wavelet transform on the discontinuous signal. By the way, a certain time series signal f
The wavelet transformed signal W (f (t)) of (t) can be expressed by the following equation. That is,

[0029]

[Equation 1] It is a function that satisfies, and is called a basic wavelet or an analyzing wavelet. In addition,
a (a> 0) is called a period parameter and b is called a shift parameter, which is a parameter to be adjusted in order to detect a discontinuity point. An example of this basic wavelet ψ (x) can be expressed by the following equation.

[0030]

[Equation 2]

The former (a) is called a Mexican hat because it resembles the shape of a hat worn by Mexicans when graphed, and the latter (b) is called a French hat.

This wavelet transform is used to find a discontinuous point or a differentiable point of the discontinuous source signal f (t), and FIGS. 2 to 6 show various time-series source signals f. The signal after wavelet transformation with respect to (t) is shown. In these figures, the horizontal axis represents time and the vertical axis represents signal magnitude.

First, in FIG. 2, 0 to 1 at time t = 0.
Original signal f (t) of a step function that changes stepwise
It is a figure which shows the relationship between the signal after that and the wavelet transformation. Since this step function cannot be differentiated at t = 0, the signal after wave transformation has a value other than 0, that is, a discontinuity point near t = 0 except t = 0.

FIG. 3 shows the ramp at t = 0.
It is a figure which shows the relationship between the original signal f (t) of the ramp function which starts to change in a circular shape, and the signal after the wavelet transformation. In this case as well, the point is near the non-differentiable point t = 0, and the signal after the wave conversion shows a value other than 0. FIG. 4 is a diagram showing the relationship between the original signal f (t) of the pulse function and its signal after wavelet transformation.

Next, FIG. 5 is a diagram showing the relationship between white noise and the signal after its wave conversion. Generally, noise is often superposed on the sensor for measuring the state of the plant, and therefore, it is detected as a very changed signal as shown in FIG. Therefore, in the case of a sensor output signal on which noise is superimposed, the signal after the wavelet transform has various slow changes, but it is difficult to determine such a sensor abnormality by catching such a changing signal.

However, as shown in FIG. 6, assuming that noise is superimposed on the measurement data of the sensor, but the output signal of the sensor temporarily falls and outputs an abnormal value, the signal after wavelet transformation is Marked changes appear in. The signal after the wavelet transformation can be clearly judged to be an abnormality of the sensor 9 or an abnormality sign.

Therefore, the data analysis means 3 in this apparatus
In 2, the measured data of the load cell 9 is wavelet-transformed as described above to find and output a discontinuous point or a differentiable point. However, even in the wavelet transform, a completely different signal waveform after the wavelet transform may be obtained depending on the time-series original signal f (t) or the values of the parameters a and b in the equation (2).

Therefore, the abnormality judging means 34 judges the abnormality of the sensor 9 from the wavelet transform signal. There are two possible methods. One of them is to prepare wavelet-transformed sensor measurement signals due to various abnormalities in advance, and compare the wavelet-transformed signals actually output from the data analysis means 32 with the previously prepared signals. , It is determined whether or not the sensor 9 is abnormal. In the determination, the actual measurement data of the sensor 9 may be held during rolling of one rolled material 1, the offline comparison may be performed after the rolling is finished, and the comparison determination result may be recorded and displayed.

Further, another abnormality judgment is to judge the presence / absence of abnormality on-line and in real time by using the high-speed processing function of the neural network. Less than,
An example using a neural network will be described. First, the sensor 9 is classified according to the rating, characteristics, and other items, and various conditions such as abnormal signals as shown in FIGS. 2 to 6 are extracted. For example, when a step-like abnormality occurs as shown in FIG. 2, when a ramp-like abnormality occurs as shown in FIG. 3, or when a pulse-like signal is finally produced as shown in FIG. Etc., and a pattern after wavelet transform of these abnormal signals is created.

Incidentally, taking the case of FIG. 6 as an example, the wavelet-transformed signal of the signal measured by the sensor 9 within the range of -2.0 to 2.0 on the horizontal axis is patterned to form an input pattern for the neural network. And FIG. 7 is an example of that.

In the figure, the vertical i and horizontal j are 11 respectively.
The "1" portion of each of the divided cells is the portion through which the signal curve after wavelet transformation passes. Therefore, the learning support means 33 gives the signal after wavelet transformation due to various abnormalities of the sensor 9 and the output value to the neural network as a teacher signal, and learns so as to obtain the optimum coupling weight coefficient.

The neural network forming the abnormality judging means 34 is as shown in FIG. That is, it is composed of an input layer, an intermediate layer, and an output layer, and the input layer is a signal obtained by wavelet-transforming the measurement data of the load cell 9 or the like, for example, "0" in 11 × 11 cells shown in FIG. Alternatively, "1" is input. Then, 11 × 11 neurons in the input layer and m neurons 1, 2, ..., M in the intermediate layer are connected to each other by a connection weight coefficient w _ijk , and each neuron in the intermediate layer and 1 in the output layer are connected. The two neurons are connected by the connection weight coefficient v _k1 .

Therefore, the learning support means 33 provides the connection weighting factors w _ijk and v _k1 with which a predetermined output value is obtained when an abnormal pattern as shown in FIG. 7 is input.
By the way, in general, learning a neural network
It is more effective if the data of the teacher signal is large to some extent. It is also known that the larger the number of neurons in each layer, the more data of the teacher signal is required. On the other hand, in the configuration of FIG. 8, if the number of neurons in the input layer is reduced, the number of teacher signals can be reduced. . Therefore, in order to reduce the number of neurons in the input layer, it is conceivable to further reduce the maximum value 11 of i and j in FIG. 7, but this causes deterioration of resolution.

Therefore, as another method, as shown in FIG. 9, it is effective to previously delete a portion (a hunting portion in FIG. 9) that cannot be taken by the signal after the wavelet transformation of the abnormal signal. In the case of FIG. 9, it is possible to reduce the number of squares 11 × 11 = 121 in FIG. 7 to 81, which is about 2/3.

Then, the abnormality judging means 34 introduces the wavelet-transformed analysis data into the neural network, and when a predetermined output value is obtained from the output layer at that time, judges that the sensor 9 is abnormal, and makes the judgment. The result is sent to the control switching means 35.

Here, the control switching means 35 switches to a method in which the sensor 9 is not used at the same time when the alarm is issued. In other words, as described above, the conventional control method of the looper can perform control without the tension value, so that the method can be switched to that method.

However, depending on the operating conditions, various control methods such as non-interference control utilizing the tension value, LQ control, ILQ control, H control, etc. may be good. Therefore, when these control methods are used, the sensor is used. The switching means 31 is switched from the sensor 9 side to the direction of calculating the tension from the current of the looper electric motor 10.

As is well known, the torque generated by the looper 8 is T _L , the torque received by the looper 8 by the tension of the rolled material 1 is T _T , the torque due to the weight of the material between stands is T _W , and the torque due to the weight of the looper is T _M. , Letting the torque of the looper 8 during acceleration / deceleration be T _A , the following relational expression holds.

T _L = T _T + T _W + T _M + T _A (6) However, in the above equation, T _L is obtained from the current of the looper motor 10, and T _W and T _M can be easily obtained from a geometrical consideration. Desired. Therefore, T _T can be easily calculated and the tension can be determined.

Therefore, according to the configuration of the above embodiment, when the tension of the continuous hot rolling mill is controlled by using the measurement data of the sensor 9, the measurement data of the sensor 9 is wavelet by the data analysis means 32. By performing the conversion, the signal is converted into a signal in which a discontinuity appears when the sensor is abnormal, so that it is possible to clearly determine whether or not the sensor 9 is abnormal in the subsequent abnormality determining unit 34. Further, the abnormality determining means 34 stores the state of the wavelet-transformed when the sensor is abnormal in advance and compares it with the wavelet-transformed signal of the actual sensor signal to automatically and quickly detect the abnormality of the sensor 9. Can be detected. Further, when it is determined that the sensor 9 is abnormal, the control automatically shifts to another control means, so that unstable control due to the abnormality of the sensor 9 can be avoided. Further, even when the sensor 9 is broken down, it can be repaired during the execution of the alternative control, and the maintenance becomes very easy.

Next, as a second embodiment of the device of the present invention,
For example, an example applied to plate thickness control between two stands of a continuous hot rolling mill will be described with reference to FIG. This control device is newly provided on the output side of the plate thickness gauge 15 for measuring the plate thickness of the rolled material 1 as compared with FIG. 1, and normally outputs the measurement data of the plate thickness gauge 15 to output the plate thickness gauge. The sheet thickness on the i + 1 stand outlet side is estimated from the sensor switching means 41 that switches to the alternative measuring means when the abnormality of 15 and the measurement data of the sheet thickness gauge 15 input via this sensor switching means 41, and the estimated sheet thickness So as to obtain the speed command value to the main machine 4 via the speed controller 6 and the rolling load P _{i of the} load cell 20 and the i stand roll gap opening S _i. Is calculated, and when the plate thickness gauge 15 is abnormal, the plate thickness control device 42 is operated through the sensor switching means 41.
And a plate thickness calculating means 43 for supplying
The structure is such that tension control is performed on one stand side.

Therefore, according to the above configuration, i +
The roll gap control device of one stand, that is, the rolling down device 19 controls the tension between the i to i + 1 stands, and adjusts the plate thickness on the delivery side of the i + 1 stand to the main unit 4 on the i stand side.
Control with.

Specifically, the plate thickness control device 42 uses the plate thickness gauge 1
The outgoing side plate thickness of the i + 1 stand is estimated from the plate thickness measurement data of the rolled material 1 according to No. 5, and the outgoing side plate thickness h of the i + 1 stand is calculated.
A speed command value of the main machine 4 is obtained so that _{i + 1} becomes a desired plate thickness, and the speed command value is supplied to the speed control device 6.

Assuming that the plate thickness measured by the plate thickness gauge 15 is h _ix , the following equation holds from the law of constant mass flow. h _{i + 1} = (v _i · B _i / v _{i + 1} · B _{i + 1} ) · H _{i + 1} (7) H _{i + 1} = h _i · e ^{-Li · s} (8) Here, v _i , v _{i + 1} : i, i + 1 stand exit side material speed (mm / s), B _i , B _{i + 1} : i, i + 1 stand exit side plate width (mm), H _{i + 1} : i + 1 Stand side plate thickness (mm), L
_i : Material transfer speed (s) from the plate thickness gauge 15 to the i + 1 stand, and e ^{−Li · s} represents delay.

Then, the measurement data of the plate thickness gauge 15 is sent to the data analysis means 32, the wavelet transform is performed as described above, and the converted signal is sent to the abnormality determination means 34. The abnormality determining means 34 learns the connection weighting coefficient while applying a teacher signal from the learning support means 33 to the neural network by using the wavelet-transformed signals of various abnormal signals of the plate thickness gauge 15. Then, depending on whether the signal after wavelet transformation by the measurement data of the plate thickness gauge 15 at the time of actual operation of the target plant is the same as the result of the wavelet transformation of the abnormal signal that has already been learned,
It is determined whether or not the plate thickness gauge 15 is abnormal. Here, when it is determined to be abnormal, the control switching unit 35 switches the plate thickness gauge 41 and supplies the gauge meter plate thickness obtained by the plate thickness calculation unit 43 to the plate thickness control device 42.

Here, the gauge meter plate thickness h _iG used in the plate thickness control device 42 of the i stand is represented by h _iG = S _i + (P _i / M _i ) ... (9) Here, S _i : i stand roll gap opening (mm), P _i : i stand rolling load (Kg), M _i :
i stand mill constant (Kg / mm 2).

Therefore, the plate thickness controller 42 delays the gauge meter plate thickness h _iG to the i + 1 stand and calculates the i + 1 stand entrance side plate thickness H _{i + 1} . H _{i + 1} = h _iG e ^{-Lis · s} (10) Here, L _is is a material transfer time (s) between stands _{i to i + 1} .

Then, using this H _{i + 1} , the plate thickness on the delivery side of the i + 1 stand is obtained from the equation (7). Therefore, according to the configuration of the above-described embodiment, the data analysis unit 32 performs the wavelet transform on the measurement data of the plate thickness gauge 15 which is a sensor to generate a signal in which a discontinuity appears when the sensor is abnormal. Since the conversion is performed, it becomes possible to clearly determine whether or not there is an abnormality in the sensor in the abnormality determining unit 34 later. The abnormality determination means 34 stores the wavelet-transformed state in advance when the sensor is abnormal, and compares it with the wavelet-transformed signal of the actual sensor signal to automatically and quickly detect the sensor plate. An abnormality of the thickness gauge 15 can be detected. Further, when it is determined that the plate thickness gauge 15 is abnormal, the control automatically shifts to another control means, so that unstable control due to the abnormality of the plate thickness gauge 15 can be avoided. Further, even when the thickness gauge 15 is out of order, it can be repaired during the execution of the alternative control, and the maintenance becomes very easy.

Next, FIG. 11 is a block diagram showing a third embodiment of the apparatus of the present invention, which is applied to, for example, plate width control of a rolled material. This device newly outputs a strip width gauge 51 for measuring the strip width of the rolled material 1, and normally outputs measurement data of the strip width gauge 51,
Board width gauge switching means 5 for switching when the board width gauge 51 is abnormal
In order to control the plate width B _{i + 1} on the delivery side of the _i + 1 stand by the tension between the i and i + 1 stands by taking in the signal relating to the plate width which is transmitted through the switch 2 and the plate width calculation means 53. sheet width control for determining the tension setting value t _fref of (AWC:
An Automatic _Width Control device 54 is provided, and the tension set value t _fref obtained here is introduced into the tension control device 14.

That is, in this apparatus, the board width control device 54 calculates the board width on the delivery side of the i + 1 stand by using the measurement data of the board width meter 51 installed between the i to i + 1 stands,
The plate width is changed to the tension set value t _fref between the i and i + 1 stands so that the plate width becomes a desired value, and the tension set value t _fref is supplied to the tension control device 14. This tension control device 14 has a tension set value t _fref.
The tension control is performed based on

Further, the measurement data of the plate width meter 51 is sent to the data analysis means 32, the wavelet transform is performed as described above, and the converted signal is sent to the abnormality determination means 34. The abnormality determining means 34 learns the connection weighting coefficient while applying a teacher signal from the learning support means 33 to the neural network using the signals obtained by wavelet transforming various abnormal signals of the plate width gauge 51. Then, depending on whether or not the signal after the wavelet-te transform by the measurement data of the strip width gauge 51 at the time of actual operation of the target plant is the same as the result of the wavelet transformation of the abnormal signal which has already been learned, the strip width gauge 51 Determine whether there is any abnormality. Here, when it is determined that there is an abnormality, the control switching unit 35 switches the plate width meter 51, and the plate width obtained by the plate width calculation unit 53 is set to the plate width control device 5.
Supply to 4.

The plate width calculating means 53 calculates a plate width which is an alternative to the plate width meter 51 using a plate width model. Therefore, also in this apparatus, similarly to the above, the abnormality of the plate width gauge 51 which is a sensor can be detected automatically and quickly. Further, when it is determined that the plate width gauge 51 is abnormal, the control automatically shifts to another control means, so that unstable control due to the abnormality of the plate width gauge 51 can be avoided. Similarly, when the plate width gauge 51 has a failure, it can be repaired during the execution of the alternative control, and the maintenance becomes very easy.

In addition, although the above embodiment has been described with reference to the continuous hot rolling mill, it goes without saying that it can be applied to the abnormality judgment of the sensors installed in other various plants. In addition, the present invention can be modified in various ways without departing from the scope of the invention.

[0064]

As described above, according to the present invention, it is possible to appropriately judge an abnormal state including a sign from the measurement data of the sensor installed in the target plant, and to automatically and quickly switch to another alternative means. It is possible to switch to secure stable control of the target plant.

Further, by using the abnormality judging means by the neural network, real-time abnormality diagnosis becomes possible, and it becomes possible to automatically judge the abnormality and switch to another alternative means, so that not only the sensor abnormality but also the maintenance and inspection of the sensor can be performed. You can continue stable control of the plant while using it easily.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of a plant control device according to the present invention.

FIG. 2 is a diagram illustrating a relationship between a step function signal and a wavelet transform.

FIG. 3 is a diagram illustrating a relationship between a ramp function signal and a wavelet transform.

FIG. 4 is a diagram illustrating a relationship between a pulse function signal and a wavelet transform.

FIG. 5 is a diagram illustrating the relationship between white noise and wavelet transform.

FIG. 6 is a diagram illustrating a relationship between a sensor abnormality and a wavelet transform in white noise.

7 is a pattern diagram of a signal after wavelet transformation when the sensor of FIG. 6 is abnormal.

FIG. 8 is a diagram showing a configuration of a neural network that constitutes an abnormality determination means.

9 is another pattern diagram of a signal after wavelet transformation when the sensor of FIG. 6 is abnormal.

FIG. 10 is a configuration diagram showing a second embodiment of the plant control device according to the present invention.

FIG. 11 shows a third plant control device according to the present invention.
FIG.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Rolled material, 2, 3 ... Rolling machine, 9 ... Load cell or strain gauge, 10 ... Looper electric motor, 14 ... Tension control device, 2
0, 21 ... Load cell, 31, 41, 52 ... Switching means,
32 ... Data analysis means, 33 ... Learning support means, 34 ... Abnormality judgment means, 35 ... Control switching means, 42 ... Plate thickness control device, 43 ... Plate thickness calculation means, 51 ... Plate width gauge, 53 ... Plate width calculation means , 54 ... Board width control device

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location G06F 15/18 8837-5L

Claims (2)

[Claims] 1. A sensor which is installed at a required location of a target plant and which measures a process state of the target plant, and data which extracts analysis data for finding discontinuity points by wavelet transform of measurement data from the sensor. Analysis means, abnormality determination means for determining whether or not the sensor is operating normally by using the analysis data obtained by the data analysis means, and a substitute for the abnormality sensor when the abnormality determination means determines an abnormality And a control switching means for switching to a control method that does not use the abnormality sensor. 2. The abnormality determining means determines a connection weighting coefficient between layers of the neural network by giving to the neural network the signal pattern after the wave transformation and the output value at various abnormal times of the sensor. While inputting the learning means and the analysis data of the data analyzing means at the time of actual operation of the target plant to the neural network, the coupling weight coefficient is set in the neural network to determine whether or not the sensor is abnormal. The plant control device according to claim 1, further comprising a determination unit.