JP2586114B2 - Method and apparatus for controlling thickness of sheet, method for monitoring correspondence between thickness distribution and thickness adjusting means, and method for manufacturing sheet - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a method and an apparatus for controlling a thickness of a sheet in a continuous sheet manufacturing process, and a correspondence between a thickness distribution and a thickness adjusting means. The present invention relates to a monitoring method and a method for manufacturing a sheet. More specifically, the means for adjusting the thickness of the discharged melt, which is provided in a large number of ferrules for discharging the melt in a sheet shape, and an accurate correspondence between the thickness distribution of the sheet material measured after molding. Thereby, the thickness and distribution of the sheet-like material can be close to the target value with high accuracy, and the thickness control method and apparatus capable of coping with the fluctuation of the sheet-like material forming process, The present invention relates to a monitoring method capable of obtaining feedback information to a sheet-like material forming process, which is effective for uniforming physical properties of a sheet-like material by displaying, and a method for manufacturing a sheet-like material using these methods.

[Prior Art] Conventionally, a melt, for example, a molten resin discharged from a die in a sheet shape is continuously formed into a sheet-like material,
For example, the sheet is passed through a stretching process, and the thickness of the sheet after passing through the sheet forming process is measured as a distribution in a sheet width direction by a thickness gauge, and the measured value is set to a preset target value. A method and an apparatus for controlling the thickness of a sheet-like material which are arranged in a width direction of the sheet-like material of a die so as to be close to each other through a control device are known. Have been.

For example, JP-A-52-36154, JP-A-52-36165
JP-A-56-120318, JP-A-56-133135, JP-A-58-78726, JP-A-60-132727, JP-A-60-225730, etc. It is mainly disclosed as a method for controlling the thickness of a resin film. In all of these conventional control methods, the correspondence between the position of the discharge melt thickness adjusting means in the die and the width direction position of the formed sheet-like material whose thickness changes when the adjusting means is adjusted is adjusted. This is a method of performing control under the assumption that it is completely taken.

However, in reality, since the above-mentioned correspondence cannot be taken accurately and accurately because of the following reasons, only a rather rough correspondence can be known, the conventional method essentially includes the sheet-like material after passing through the sheet-like material forming process. There is a problem that the thickness control accuracy is limited.

For example, FIG. 13 conceptually shows a process of forming a conventional biaxially stretched plastic film as a sheet.
The molten resin 2 discharged from the die 1 in a sheet shape is biaxially stretched by a longitudinal stretching process (also referred to as a longitudinal stretching process) 3 and a width stretching process (also referred to as a transverse stretching process) 4, and has a predetermined thickness. As a wide biaxially stretched plastic film 5 having The thickness distribution of the plastic film 5 in the width direction is measured by, for example, a scanning thickness gauge 6. When the width of the plastic film 5 becomes wide, for example, about 5 to 10 m, the measuring point by the thickness gauge 6 becomes, for example, 30 in the film width direction.
As a result, the number of discharge melt thickness adjusting means disposed on the base 1, for example, the number of the adjusting bolts 7 for adjusting the slit gap of the base discharge port becomes about 100. Correspondence between the positions of the adjustment bolts 7 arranged in such a large number and the positions where the thickness of the film 5 changes when the adjustment bolts 7 are adjusted, that is, the thickness of the adjustment bolts 7 after forming the film The following compulsory tests are usually performed to measure and understand the degree of impact on As shown in Fig. 14, any appropriate extracted adjustment bolt 7 is forcibly shaken (for example, about 20% of the allowable adjustment amount), and the film thickness distribution position 8 that changes particularly greatly at that time is measured. Therefore, there is a method of taking correspondence between the two. The center position in the width direction of the plastic film 5 after molding and the center position in the width direction of the die adjusting bolts 7 are the positions of the ears at both ends in the width direction of the plastic film 5 and the both ends in the width direction of the die. There is also a method of simply assigning each adjustment bolt position and the position in the width direction of the plastic film 5 on the assumption that the adjustment bolts 7 correspond exactly. Further, in this method, there is also known a method of assigning masses in consideration of the fact that the positions of the ears at both ends of the plastic film 5 are thicker than other portions.

However, in order to grasp the correspondence between the position of the adjustment bolt 7 and the position of the change in the thickness of the film 5 by the method shown in FIG. It is necessary to forcibly generate large thickness unevenness that can be clearly recognized. The film 5 having such a large uneven thickness is of course not a product, and a test for grasping the above correspondence is the same as a kind of destructive inspection. Therefore, a large amount of loss occurs during the test, and it takes several hours to restore the forcibly changed adjustment bolts 7 to their original positions and to keep the product thickness level and its distribution within the range that can be commercialized. And the production efficiency is greatly reduced.

In addition, a method of obtaining the correspondence between each adjustment bolt 7 and the position in the film width direction by uniform or mass allocation based on the above-mentioned certain adjustment bolt 7 or the forced test shown in FIG. 14 at appropriate intervals. In a method in which a proper number of adjustment bolts 7 are set, the correspondence between the adjustment bolts 7 in between and the position in the film width direction are equally allocated, and determined by mass allocation, the film thickness by the base 1 and the thickness gauge 6 is used. Since a stretching process, particularly a width stretching process 4 is interposed between the measuring position and the stretching in this portion, it is not always possible to perform completely uniform stretching. It is almost impossible to take action. Further, since the film forming processes 3 and 4 cannot always be completely constant conditions in relation to the film itself or the film to be formed, processing and stretching conditions slightly vary. Along with this variation, the correspondence between the position of each die adjusting bolt 7 and the width direction position of the formed film 5 also changes as shown by the arrow 9 in FIG. It is difficult to grasp. Since such an accurate response has not been taken in reality, even if the thickness gauge 6 can accurately measure the thickness, the signal is fed back to control the adjustment bolts 7. In the stage, the inappropriate adjustment bolt 7 is controlled, and eventually, there is a limit that the thickness control level of the film 5 cannot be broken.

In addition, there is the following problem with respect to the control adjustment amount of the adjustment bolt 7. In other words, for example, by a forced test as shown in FIG. 14, when a certain adjusting bolt 7 is greatly changed (for example, about 20%), the amount of change in thickness at a certain width direction position of the formed film 5 can be measured. However, the adjustment amount of the adjustment bolt 7 to be changed in the actual thickness control during production is as small as 0.1 to 0.5%, for example. It is not always the case that it is proportional to the relationship obtained in the above compulsory test. This is mainly due to the presence of the molding processes 3 and 4 between the base 1 and the thickness gauge 6, and the relationship between the adjustment amount of the adjustment bolt 7 and the thickness change amount of the film 5 after molding is also complicated. It is a correspondence. Furthermore, if the processes 3 and 4 are conditionally fluctuated, it is impossible to obtain a uniquely accurate response.

Further, if the correspondence relationship between the width direction position of the die adjusting bolt 7 and the width direction position of the formed film 5 is added, the following problem occurs.

As a result of a very precise test performed by the present inventors, when the adjustment bolt 7 with the base 1 is adjusted by a small amount of one unit operation (for example, about 0.1 to 0.5% of the allowable adjustment amount), it corresponds to the adjustment. The thickness change pattern of the film 5 at the width direction position of the film 5 after molding (this correspondence can be obtained in advance by a forced test as shown in FIG. 14) is, for example, as shown in FIG. It turned out to be a pattern. In other words, the unit operation of a certain adjustment bolt 7 changes like the thickness profile 10 portion around the center position of the position of the film 5 corresponding to the bolt 7. At the same time, the thickness profile 11a, 1
It changes in the opposite direction to the profile 10 as shown in 1b. This is because when a certain portion in the width direction of the die 1 is locally thickened, for example, the molten material on both sides is drawn in by the influence of the flow of the molten material at that portion, and these changes occur after molding. It appears that the pattern appears as a thickness profile pattern. As described above, since the central portion 10 changes a certain thickness with respect to the unit operation amount of the adjusting bolt 7, and both side portions 11a and 11b change in the opposite direction to the central portion 10, as described above, each adjusting bolt is changed. If the position of the base 1 in the width direction of the die 7 and the position of the width direction of the film 5 after molding are not accurately correlated, the place where the film thickness should be corrected is corrected to be thinner, or the film thickness is corrected. Doing so would lead to an unsatisfactory limit on the thickness control level.

[Problems to be Solved by the Invention] As described above, an object of the present invention is to provide a conventional thickness control method in which the position of a melt thickness control means of a die and a sheet-like material after passing through a sheet-like material forming process. In view of the fact that there is a limit that the thickness adjustment level cannot be satisfied as an essential factor that the correspondence relationship with the width direction position cannot be accurately grasped, the above-mentioned correspondence relationship is subjected to a forced test that greatly changes the adjustment bolt as described above. Without performing, it can be obtained with much higher accuracy than before, and by determining the control position and amount of the discharge melt thickness adjustment means of the die based on the correspondence, the thickness control accuracy of the sheet-like material can be It is an object to provide a method and a device which can be greatly improved.

Another object of the present invention is to provide a method and an apparatus for controlling a thickness of a sheet-like material capable of automatically coping with a variation in the sheet-like material forming process even when there is a variation in the process. Is to provide.

Another object of the present invention is to use a part of the steps of the above-described thickness control method to make the physical properties of the sheet-like material uniform in the width direction from the data in the width-direction thickness adjustment of the sheet-like material. It is to obtain feedback information to an effective sheet forming process.

Further, another object of the present invention is to manufacture a sheet having a thickness controlled with high accuracy by using the above-described thickness control method.

Means for Solving the Problems In order to achieve the above object, a method for controlling the thickness of a sheet according to the present invention is configured as follows.

That is, the molten material discharged in a sheet form from the die is continuously passed through a sheet forming process, and the thickness of the sheet after passing the sheet forming process is measured as a distribution in a sheet width direction. Then, a sheet-like object that controls a plurality of discharge melt thickness adjusting means disposed in the sheet-like object width direction of the die via a control device so that the measured value approaches a preset target value. In the thickness control method, a Kalman filter is incorporated in the control device, and the position and adjustment amount of the die of the discharge melt thickness adjusting means in the sheet width direction and the adjustment amount of the sheet by using the Kalman filter and the sheet forming process. The correspondence relationship between the position where the thickness of the sheet-like material changes in the width direction of the sheet-like material after passing and the thickness of the sheet-like material by the adjustment of the discharge melt thickness adjusting means and the amount of the change is shown. As a relational expression including a Rix, an estimation is performed while sequentially correcting the correspondence matrix so that the actual thickness distribution measured value in the sheet-like object width direction sequentially input approaches the target value, and the successive estimation An optimum output to the discharge melt thickness adjusting means is determined based on the obtained relational expression including the correspondence relationship matrix, and an adjustment margin of each discharge melt thickness adjusting means is controlled based on the optimum output. This is a characteristic sheet thickness control method.

Further, the thickness control method for a sheet-like material according to the present invention continuously passes a molten material discharged from a die into a sheet-like shape through a sheet-like material forming process, and forms a sheet-like material after passing through the sheet-like material forming process. The thickness of the object is measured as a distribution in the width direction of the sheet-like object, and a large number of the bases are arranged in the width direction of the sheet-like object via a control device so that the measured value approaches a preset target value. A method of controlling the thickness of a sheet-like material for controlling a discharge melt thickness adjusting means, comprising: (1) (a) (a) a position and a position of an adjustment means in a discharge melt width direction; (B) the position of the sheet-like material whose thickness changes due to the adjustment of the adjusting means in the sheet-width direction after the sheet-like material is formed by the sheet-like material forming process; And the time following said time Estimating the correspondence relationship of the thickness change amount, and in obtains the correspondence matrix (A * _^i-1) with respect to the respective adjusting means by its estimate, (ii) certain sheet at time (t _i-1) measuring the width direction thickness distribution of the object (P _i-1), was measured (2) widthwise thickness distribution of the sheet in the next time _{(t i) (P i)} , (3) the estimated Using an equation including the calculated correspondence matrix (A ^* _i-1 ) and the measured thickness distribution (P _i-1 ), ( _i ) the estimated thickness distribution (P ^* _i ) at the time (t _i ) )
And calculating the difference between the estimated thickness distribution (P ^* _i ) and the measured thickness distribution (P _i ). (Ii) If there is the difference, the estimated correspondence matrix ( A ^* _i-1 ) is corrected to obtain a corrected estimated correspondence matrix (A ^* _i ), and the above steps (i) and (ii) are performed at each time (t
_{i + 1} , t _{i + 2,} etc.) and the estimated correspondence matrix (A ^* _i , A ^* _{i + 1,} etc.) and the measured thickness distribution (P _i , P
_{i + 1,} etc.), and repeatedly execute the Kalman filter to obtain the measured thickness distribution (P _{i + 1} ,
P _{i + 2,} etc.) target thickness distribution () the estimated relationship matrix to have been the following modifications closer _{^{(A * i + 1, A}} * i + 2
And (4) calculating the optimum adjustment amount of each adjusting means according to the target thickness distribution () from the estimated correspondence matrix obtained in step (3). Automatically controlling the amount of adjustment of the adjusting means to obtain an optimal thickness distribution of the sheet based on the amount.

Further, the sheet thickness control device according to the present invention,
A die for discharging the molten material into a sheet, a molding process device for molding the sheet discharged from the die, and a distribution of the thickness of the sheet formed by the molding process in the width direction of the sheet. A sheet thickness control device comprising: a sheet thickness measuring means for measuring as; and a discharge melt thickness adjusting means disposed in a large number in a width direction of the die in the sheet material, The control device has a Kalman filter, and by using the Kalman filter, the position of the discharge melt thickness adjusting means in the width direction of the die and the adjustment amount of the die and the sheet after passing through the molding process device. In the width direction, the relationship between the position where the thickness of the sheet-like material changes due to the adjustment of the discharge melt thickness adjusting means and the amount of change is defined as a relational expression including a correspondence relation matrix. The estimation is performed while sequentially correcting the correspondence matrix so that the actual thickness distribution measured value in the sheet width direction sequentially inputted approaches a preset target value, and the correspondence obtained by the successive estimation is estimated. Based on a relational expression including a relation matrix, an optimum output to the discharge melt thickness adjusting means is determined, and an adjustment margin of each discharge melt thickness adjusting means is controlled based on the optimum output. What you do.

The above Kalman filter was published by REKalman in 1960. When there are multi-dimensional fluctuation factors (forced terms) in the system equation, actual data (observed values) that are sequentially input with the multi-dimensional fluctuation factors , And the next observation value is estimated using the multidimensional variation factor term that is successively estimated and corrected using the actual observation value.

Here, the thickness distribution measurement value (thickness profile measurement value) of the sheet-like material after passing through the sheet-like material forming process and the adjustment amount of each discharge melt thickness adjusting means of the die are discretely with respect to time. Since it is a continuous value, any continuous time is t
_i-1 , t _i , t _{i + 1} , where t _i is the current time, the thickness distribution measurement values of the sheet-like material at time t _i-1 and time t _i are P _i-1 and P _i , respectively. in stands, further represents the difference between the measured value of the thickness distribution of the sheet at time t _i-1 and time t _i in [Delta] P _i, is _{ΔP i = P i -P i-} 1 (1).

When the adjustment amount of each adjusting means of the base at time t _i-1 is denoted by B _i-1 , the measured value of the thickness distribution of the sheet-like material at time t _i
P _i is the result of B _i−1 , that is, at the current time t _i , B _i−1 ,
_Pi-1 and _Pi are all known quantities.

In the present invention, if the correspondence matrix of the sheet-like material forming process at time t _i-1 is represented by A _i−1 , the system equation of the sheet-like material forming process is ΔP _i = A _i−1 * B _i−1 + Ε _i (2). Here, ε _i is a noise (error) vector.

As described above, for example, m =
Since it can be represented by a vector consisting of about 300 discrete values, and the discharge melt thickness adjusting means can be represented by a vector consisting of, for example, n = about 100 discrete values (corresponding to, for example, the number of the aforementioned die adjusting bolts), The above system equation is Can be rewritten. That is, A _i-1 is a multidimensional correspondence matrix between ΔP _i and B _i-1, and the multidimensional correspondence matrix A _i-1 is the thickness of the actual sheet-like object sequentially input by the Kalman filter. It is sequentially estimated while being successively corrected using the measured values of the distribution, that is,
A _i-1 makes it possible to grasp with high accuracy the correspondence between the means for adjusting the thickness of the melt to be discharged from the die and the thickness distribution of the formed sheet.

Here, assuming that the optimal correspondence matrix estimated at the time t _i-1 is A ^* _i-1 , the following relationship is established from the equation (2).

^{_{ΔP * i = A * i-}} 1 * B i-1 + ε i (4) ΔP * i is the sheet at time t _i-1 and the current time t _i which is calculated by using the ^{A *} _i-1 This is an estimated value of the difference in thickness distribution, and as a result, the following equation is derived from equation (1).

^{_{P * i = P i-1}} + ΔP * i = P i-1 + A * i-1 * B i-1 + ε i
(5) P ^* _i is the measured value P _i of the thickness distribution of the sheet at the present time t _i
Is an estimated value of the thickness distribution of the sheet-like object at the current time t _i before is obtained.

In the Kalman filter, the current time t calculated using the measured value P _i of the thickness distribution of the sheet-like object obtained at the current time t _i and the optimal correspondence matrix A ^* _i-1 estimated at the time t _i-1.
The optimum correspondence matrix A ^* estimated at time t _i-1 based on the difference between the _i and the estimated value P ^* _i of the thickness distribution of the sheet-like material ^.
_i-1 is corrected and the optimal correspondence matrix A for the current time t _i
^* Estimate _i .

The variables multidimensional referred to in the present invention, the current time t _i to the measured value P _i of the thickness distribution of the sheet was calculated using the estimated value A ^{* _i-1} at the current time t _i Sheet The noise is removed from the difference from the estimated value P ^* _i of the thickness distribution of the object.

Further, it is possible to calculate the optimum adjustment amount B ^* _i of the die melt thickness adjusting means at the time t _i . Specifically, the current time
Using the optimal correspondence matrix A ^* _i estimated at t _i , the following relationship is established from equation (5).

P ^* _{i + 1} = _Pi + A ^* _i * _Bi + εi _{+ 1} (6) _Pi and A ^* _i are known quantities, and P ^* _{i + 1} is the melt melt of the mouthpiece operated at time t _i. It is an estimated value of the thickness distribution of the sheet-like object at the time t _{i + 1} with respect to the adjustment amount B _i of the thickness adjusting means.

If the thickness distribution where the target sheet, the calculation of B ^* _i which the difference of the P ^{* i} _{+ 1} for B _i to a minimum can be performed as follows.

That is, the evaluation function J is By calculating B _i that minimizes J, B ^* _i can be obtained. Thus (7) may be obtained a B _i that satisfies the extremes of expression.

dJ / dB _i = 0 (8) B _i that satisfies the expression (8) is the optimum adjustment amount B ^* _i of the die melt thickness adjusting means at the time t _i . In other words, by estimating the optimal control amount, a highly accurate thickness control level that cannot be obtained by the conventional method is achieved.

In the above-described successive estimation, the data is fed back every time the thickness distribution of the sheet is actually measured, and the correspondence matrix A is sequentially corrected using the data and the data up to that time. Therefore, even if a change occurs in the sheet-like material forming process, the optimum correspondence matrix A is continuously obtained by automatically following the change, and highly accurate thickness control is ensured.

That is, in the statistical sequential estimation using the Kalman filter, the thickness distribution of the sheet after passing through the sheet forming process and the control position and the control amount of the discharge melt thickness adjusting means of the die are adjusted. The control device itself learns itself so that an accurate response can be taken, and automatically corrects it using actual observation data.

However, since the process of obtaining the correspondence matrix is a sequential estimation, it is necessary to set a certain initial value. An appropriate initial value may be used. For example, the state at the end of the previous production (correspondence relationship between the thickness distribution of the sheet-like material and the thickness control means for the melt to be ejected) or a rough correspondence relationship by the forced test as shown in FIG. Sufficiently, using the Kalman filter as described above as a starting point, estimation is rapidly and successively performed so that an accurate correspondence matrix can be obtained.

Next, in the method for monitoring the correspondence between the thickness distribution of the sheet-like material and the thickness adjusting means according to the present invention, the melt discharged in the form of a sheet from the die is formed by passing the molten material through a sheet-like molding press. Correspondence relationship between the thickness distribution in the width direction of the sheet-like material and the adjusting means for adjusting the thickness of the discharge melt that is disposed in the width direction of the discharge melt discharged from the die in a sheet shape. (1) For each adjusting means, (i) (a) the position of the adjusting means in the discharge melt width direction and the adjustment amount of the adjusting means at a certain time; and (b) the sheet-shaped molding Correspondence relationship between the position of the sheet material whose thickness changes due to the adjustment of the adjusting means in the width direction of the sheet material after the sheet material is formed by the process, and the thickness change amount at the time following the time. Is estimated. A correspondence matrix (A ^* _i-1 ) is obtained for all adjustment means, and (ii) a thickness-direction thickness distribution ( _Pi-1 ) of the sheet at a certain time (ti _-1 ) is measured. 2) The thickness distribution (P _i ) in the width direction of the sheet at the next time (t _i ) is measured. (3) The estimated correspondence matrix (A ^* _i-1 ) and the measured thickness Using the equation including the distribution (P _i-1 ), ( _i ) the estimated thickness distribution (P ^* _i ) at the time (t _i )
And calculating the difference between the estimated thickness distribution (P ^* _i ) and the measured thickness distribution (P _i ). (Ii) If there is the difference, the estimated correspondence matrix ( A ^* _i-1 ) is corrected to obtain a corrected estimated correspondence matrix (A ^* _i ), and the above steps (i) and (ii) are performed at each time (t
_{i + 1} , t _{i + 2,} etc.) and the estimated correspondence matrix (A ^* _i , A ^* _{i + 1,} etc.) and the measured thickness distribution (P _i , P ^* _{i + 1), respectively.}
) To obtain a modified estimated correspondence matrix (A ^* _{i + 1} , A ^* _{i + 2,} etc.) at subsequent times using the Kalman filter. Estimated correspondence matrix (A ^*
_{i + 1} , A ^* _{i + 2,} etc.) and at least the position of the adjusting means in the discharge melt width direction and the width of the sheet after passing through the sheet forming process in which the thickness changes by adjusting the adjusting means. Displaying the correspondence with the position in the direction on the display means.

In this monitoring method, a step of obtaining a correspondence matrix using a Kalman filter in the above-described thickness control method is used, and from the obtained correspondence matrix, it is effective to uniform the physical properties of the sheet-like material in the width direction. Feedback information to the sheet forming process is obtained.

In this monitoring method, not all the steps of the thickness control method according to the present invention described above are necessarily required. To calculate the correspondence matrix using the Kalman filter, the thickness distribution in the width direction of the sheet-like material after the process of the sheet-like material obtained as a result of the adjustment with the adjustment amount data of each discharge melt thickness adjusting means is performed. It is only necessary that the data is repeatedly input. Therefore, the adjustment of the discharge melt thickness adjusting means based on the sheet thickness distribution data may be controlled via a conventionally known control device, or may be simply performed manually. However, in these cases, the level of control accuracy of the thickness distribution of the sheet after passing through the sheet forming process, that is, the difference between the target value and the actually appearing thickness distribution remains at the conventional level, and the sheet forming process is not performed. Only feedback information to is obtained. When both the thickness control method and the monitoring method according to the present invention are applied, the thickness distribution of the sheet is controlled with extremely high accuracy, and at the same time, feedback information to the sheet forming process is obtained.

Feedback information to the sheet forming process, which is effective for equalizing the physical properties of the sheet in the width direction, is displayed on display means, for example, a display device or a printer. The display means includes, at least, a position of the discharge melt thickness adjusting means in the sheet width direction and a sheet-like material having passed through a sheet-like material forming process in which the thickness changes by adjusting the discharge melt thickness adjusting means. The corresponding relationship with the position in the width direction is displayed.

The following information can be obtained from this correspondence. The discharge melt thickness adjusting means is usually arranged at a uniform pitch in the width direction of the sheet of the die. However, in the sheet-like material after passing through the sheet-like material forming process, the pitch between the width-wise positions of the sheet-like material corresponding to the position of the discharge melt thickness adjusting means due to variations in various conditions in the sheet-like material forming process. Are not always uniform. When the non-uniformity of the pitch is known, it is often possible to determine how to correct the temperature conditions, mechanical conditions, and the like in the sheet forming process in order to make the pitch uniform. Therefore, by making appropriate corrections to the temperature conditions, mechanical conditions, and the like, the pitch approaches the uniform direction. The fact that the pitch between the positions in the width direction of the sheet corresponding to the position of the melt thickness controlling means is uniform indicates that the processing in the sheet forming process has been performed uniformly. Therefore, by making the pitch uniform, the physical properties of the obtained sheet material are made uniform in the width direction.

Further, in the present invention, a sheet-like material having extremely excellent thickness uniformity is manufactured by using the above-described method for controlling the thickness of the sheet-like material.

That is, in the method for producing a sheet according to the present invention, the molten material is discharged from gold into a sheet and continuously passed through a sheet forming process to form a sheet, and the sheet is passed through the sheet forming process. The thickness of the subsequent sheet-shaped object is measured as a distribution in the sheet-shaped object width direction, and a large number of the bases in the sheet-shaped object width direction via the control device are controlled so that the measured value approaches a preset target value. In the method for manufacturing a sheet-like material for manufacturing a sheet-like material while controlling the disposed discharge melt thickness adjusting means, a Kalman filter is incorporated in the control device, and the discharge melt thickness adjustment is performed using the Kalman filter. The position of the die of the means in the sheet width direction and the adjustment amount thereof and the adjustment of the discharge melt thickness adjustment means in the width direction of the sheet after passing through the sheet forming process are adjusted. The relationship between the position at which the thickness of the sheet-like material changes and the amount of the change is defined as a relational expression including a correspondence matrix, and the actual thickness distribution measurement value in the sheet-like material width direction that is sequentially input is the same. Estimating the correspondence matrix while sequentially correcting it so as to approach the target value, and determining an optimum output to the discharge melt thickness adjusting means based on a relational expression including the correspondence matrix obtained by the successive estimation. And manufacturing a sheet-like material while controlling an adjustment margin of each discharge melt thickness adjusting means based on the optimum output.

[Example] Hereinafter, a specific example of the present invention will be described using a biaxially stretched plastic film manufacturing process as an example.

FIG. 1 shows a biaxially stretched plastic film manufacturing process in which a control process for implementing the method and apparatus of the present invention is applied.

In the figure, 21 is a sheet of molten resin as a melt.
Reference numeral 22 denotes a discharge base. The molten resin 22 discharged in a sheet shape is cooled and formed by a cooling drum 23, and then passes through a longitudinal stretching process (longitudinal stretching process) 24 having a preheating, longitudinal stretching, and heat fixing (cooling) process. It is stretched in the longitudinal direction, and then stretched in the width direction by passing through a width stretching process (transverse stretching process) 25 having a preheating, width stretching, and heat setting (cooling) process, and is formed into a film 26 having a predetermined thickness. After being wound, it is continuously wound on a suitable winder 27.

The sheet forming process in the present invention means all of the cooling drum 23, the longitudinal stretching process 24, and the width stretching process 25. However, since the present embodiment is a two-width stretching process, the above-described process is employed. However, another sheet-like material forming process, for example, a sheet-like material forming process having a process of cooling and molding a discharged molten resin and a longitudinal stretching process is used. It may be a process (uniaxial stretching process).
Further, a device having a stretching process after the biaxial stretching process may be used.

In the present embodiment, after the cooling drum 23, after the longitudinal stretching process 24, and after the width stretching process 25, respectively, thickness gauges 28, 29, 30 for measuring the thickness of the film as a distribution in the width direction.
Is installed. The measurement data obtained by the thickness gauges 28, 29, and 30 are input to a CPU 31 as a control device. In the CPU 31, at the position of each thickness gauge 28, 29, 30
A desired target thickness distribution in the film width direction can be input and set.

The thickness of the molten resin 22 to be discharged can be adjusted in a direction perpendicular to the paper surface of FIG. 1, that is, in the sheet width direction of the discharged molten resin in the form of a die. A large number of, for example, about 100 adjustment bolts 32 capable of adjusting the slit of the base 21 as means are arranged at a suitably fine pitch. The structure of this part is, for example, as shown in FIG. 2, and an adjustment bolt is provided on one side block 34 forming the slit-shaped discharge port 33 of the base 21.
32 are provided in large numbers in a direction perpendicular to the paper surface of FIG. A heater 35 is embedded or juxtaposed in the adjustment bolt 32, and the Y of the adjustment bolt 32 is controlled by controlling the temperature of the heater 35.
The amount of expansion and contraction in the −Y direction is adjusted, whereby the slit gap of the slit-shaped discharge port 33 can be adjusted to adjust the thickness of the discharge sheet. In the present embodiment, the temperature control by the heater 35 is performed by controlling the ON rate of the heater 35. For example, as shown in FIG. 3, the ON / OFF of the heater 35 is controlled at a pitch of 10 seconds. 5 as shown
If the heater 35 is turned on for a second, the ON rate is controlled to be 50%. The control of the ON rate can be controlled even with a considerably small adjustment amount, and for example, it is possible to easily change the heater ON time by 0.1 second or the like.

The heater 35 may be controlled by controlling the applied voltage. In addition, as a discharge melt thickness adjusting means, an adjusting bolt having no heater is used, and a motor that can rotationally drive the adjusting bolt is connected to the adjusting bolt, and the rotation amount of the adjusting bolt is controlled by the motor. Is also good.

A Kalman filter is incorporated in the CPU 31. In the Kalman filter, the method of finding the correspondence matrix A shown in the above equation (5) is performed as conceptually shown in FIG. However, in this embodiment, since three thickness gauges 28, 29, and 30 are installed at each position, the thickness distribution of the film after each process and the adjustment bolts of the base 21 are adjusted.
It is possible to take correspondence with 32. It is possible to control the film thickness for each process, and to perform more precise control to bring the film thickness and the thickness distribution after the final process closer to the target values with higher accuracy.

The successive estimation shown in FIG. 4 will be described as that at the position of the thickness gauge 30, that is, in the thickness distribution control at the stage of forming into a predetermined product shape after passing through the sheet forming process. From the thickness gauge 30, the thickness distribution as the actual measurement data is input to the CPU 31 discretely every time, such as at time t _i and the next time t _{i + 1} . In the Kalman filter of the CPU 31, at time t _i , the thickness distribution at time t _i that would appear as a result when the adjustment of a certain adjustment bolt 32 was performed at time t _i-1 using the measurement data before that was also used. The estimated value P ^* _i is estimated by Expression (5) including the optimal estimated value A ^* _i-1 of the correspondence matrix. This estimated value is compared with the actual measurement data at time t _i , and is calculated as a deviation distribution (deviation profile).
An optimal correspondence matrix A ^* while correcting the deviation distribution with Kalman gain so that the deviation distribution approaches zero ^.
Estimate _i . This calculation process is repeatedly performed each time actual data is input from the thickness gauge 30. As a result, the deviation distribution gradually approaches 0, and therefore, the correspondence matrix A is most appropriately adjusted with the adjustment bolt 32 and the thickness distribution. Is estimated as an optimal matrix while being sequentially updated. However, since it is a sequential estimation, it is necessary to input only an appropriate initial value as described above.

Using the correspondence matrix A successively estimated, as shown in the above equation (7), the adjustment bolt 32 is used so that the data measured by the thickness gauge 30 approaches the target value preset in the CPU 31. Can be calculated. By the output command based on this calculation, the ON rate of the heater 35 of each adjustment bolt 32 is controlled.

In this control, for example, as shown in FIG.
By the adjustment control of a certain adjustment bolt 32, the thickness distribution of the film at the corresponding position is minutely changed in the shape as shown in FIG. , And these minute changes appear in a laminated form in the film width direction. Then, the correspondence matrix A is sequentially estimated so that the profile in the entire width direction becomes a preset target profile 42 as shown in FIG. 6, for example.
Is used to control each adjustment bolt 32.

As described above, in the method of the present invention using the Kalman filter, since the correspondence matrix A is obtained with high accuracy by statistical successive estimation, the position of each adjustment bolt 32 and the width direction position of the formed film are determined. And the degree of influence of each adjustment bolt 32 on the film thickness distribution per minute unit operation is accurately grasped. Then, the thickness profile (thickness distribution) that actually appears as a result is controlled so as to approach the target value with high accuracy.

In addition, since the actual measurement data of the thickness gauge 30 is input to the CPU 31 for each measurement, and the correspondence matrix A is sequentially corrected based on the data, the optimal matrix is obtained in a short time, and the process varies. In such a case, the automatic following is automatically performed, and the optimal correspondence matrix A is sequentially updated according to the fluctuation. Therefore, the above-mentioned optimal correspondence is always maintained during the production.

Although the above description has been made for the position of the thickness gauge 30,
Similar control is performed for the positions of the thickness gauges 28 and 29, and the positions of the thickness gauges 28, 29, and 30 are controlled, so that more accurate association can be achieved.

Further, the present invention is not limited to resin, but may be any resin having a sheet forming process such as glass and paper.

Next, FIGS. 7 to 10 show an embodiment of a method for monitoring the correspondence between the thickness distribution of the sheet-like material and the thickness adjusting means according to the present invention.

In the system shown in FIG. 7, display means is added to the system shown in FIG. A display device 51 as a display means 50 and a printer 52 are connected to the CPU 31. In the present embodiment, both the display device 51 and the printer 52 are provided, but either one may be provided.

The CPU 31 has an arithmetic unit 5 incorporating a Kalman filter.
3 are built in. The arithmetic unit 53 uses the Kalman filter to obtain the optimum correspondence matrix A ^* and the film thickness from the data on the thickness distribution of the film from the thickness gauges 28, 29 and 30 and the data on the adjustment amount of each adjustment bolt 32. finally adjusting the amount of the adjustment bolt 32 that a uniform distribution B ^*
Is calculated.

The correspondence matrix A ^* is displayed on the display means 50. The correspondence matrix A ^* has, for example, contents as shown in FIG. In FIG. 8, the sequence M ₁ , M ₂ ,
M ₃ ... Represent the amount of change in the thickness of the film after passing through the sheet forming process and the position of the thickness change in the film width direction when each of the adjustment bolts 32 is adjusted by the unit operation amount. This sequence is a film thickness variation pattern L ₁ ,
L ₂ , L ₃ ... This correspondence matrix A ^* is displayed on the display means 50 in the graph of FIG. 9, where each row corresponds to the adjustment bolt 32, and the horizontal axis thereof indicates the thickness measurement position in the width direction of the film after passing through the sheet forming process. The vertical axis indicates the amount of change in film thickness when each adjustment bolt 32 is operated by a unit operation amount. Each position F ₁ ,
.. Indicate positions corresponding to the adjustment bolts Nos. 1, ₂ , ₃ ,.

The pitch D between the positions F ₁ , F ₂ , F ₃ ... Varies in various ways depending on the conditions of the sheet forming process. Therefore, by observing the variation of the pitch D,
Variations in processing conditions in the sheet forming process can be estimated. For example, in the case as shown in FIG. 9, the Q part has a large degree of stretching in the width direction, but the R part has a small degree of stretching. From this information, to make the pitch D uniform,
A method of correcting temperature conditions, mechanical conditions, and the like in the sheet forming process can be estimated. When these conditions are appropriately corrected, the pitch D becomes uniform, the treatment in the sheet-like material forming process becomes uniform, and a film having uniform physical properties in the width direction is obtained.

The above information is obtained for each position of the thickness gauges 28, 29, 30. FIG. 10 shows a correspondence matrix at each position of the thickness gauge displayed on the display means 50, which is synthesized and simultaneously displayed. In the example shown in FIG. 10, the correspondence matrix ¹ A ^* is obtained having a pitch D ₁ relatively uniform in position in the thickness gauge 28. This indicates that relatively uniform processing was performed in the cooling forming process 23. The position of the thickness gauge 29, has a relationship matrix ² A ^* with variations in the pitch D _2, shows that the longitudinal stretching process 24 there is a variation of some process conditions. The position of the thickness gauge 30, and has a relationship matrix ³ A ^* that the pitch D ₃ varies quite large. This is because the film goes through a width stretching process 25,
Variations in processing conditions in the stretching process 25
According to directly greatly affects the variation of D _3. As described above, it is possible to grasp the variation in the conditions for each sheet-like material forming process, and to perform an appropriate action for each process.

The display means 50 displays the film thickness variation H (shown in FIG. 9) when each adjustment bolt 32 is operated by a unit operation amount, in addition to the correspondence of the positions, so that the effectiveness of each adjustment bolt is You can understand the condition. When the display content is used only as feedback information to the sheet-like material forming process as described above, only the correspondence between the position of the adjustment bolt 32 and the position in the film width direction may be displayed.

The monitoring method according to the present invention can be applied to the case where the film thickness control is performed by a conventional system.

 11 and 12 show examples.

In the system shown in FIG. 11, a Kalman filter is not incorporated in the CPU 60, but a Marman filter is incorporated only in the CPU 61 as an arithmetic unit connected to the CPU 60. Via the CPU 60, the data of the adjustment amount of the adjustment bolt 32 and the thickness distribution data measured by the thickness gauges 28, 29, 30 are repeatedly input to the CPU 61. Even when this repetitive input data is used, the correspondence matrix A ^* is obtained by the same operation as described above. The determined A ^* is displayed on the display device.

In the system shown in FIG. 12, the adjustment bolt 71 is manually adjusted while checking the thickness distribution data from the thickness gauges 28, 29, and 30. The adjustment amount of the adjustment bolt 71 is input to the CPU 72 as an arithmetic unit via an appropriate detecting means, and the data of the film thickness distribution appearing as a result of the adjustment of the adjustment bolt 71 is output from each of the thickness meters 28, 29, 30 to the CPU 72. Is input to This input is repeated, and the CUP 72 calculates the correspondence matrix A ^* using the Kalman filter. A calculated
^* Is displayed on the display device 73.

Also in the system shown in FIGS. 11 and 12, the correspondence matrix A ^* is obtained, and the display thereof provides effective feedback information for the sheet forming process. However, in these cases, since the actual film thickness control is performed by the conventional method, it is not expected that the thickness distribution is improved more than the conventional level. However, it is possible to indirectly improve the thickness distribution level by appropriately modifying the conditions of the sheet forming process based on the feedback information.

It is needless to say that the monitoring method of the present invention is not limited to the process of manufacturing a resin sheet, but may be applied to any manufacturing process having a process of forming a sheet such as glass and paper.

[Effects of the Invention] As described above, when the method and the apparatus for controlling the thickness of a sheet-like material according to the present invention are used, the thickness of the sheet-like material after molding is adjusted using a Kalman filter. Since the correspondence with the distribution is grasped with much higher accuracy than the conventional method, and the thickness of the sheet is controlled based on it, in order to make the thickness distribution of the sheet close to the target value In addition, it is possible to accurately control the thickness control means of the melt to be discharged, and it is possible to significantly improve the accuracy of controlling the thickness of the sheet-like material.

In addition, even when there is a variation in the sheet forming process, the variation can be automatically followed to maintain the high precision thickness control accuracy during production.

Further, when the method for monitoring the correspondence between the thickness distribution of the sheet and the thickness adjusting means of the present invention is used, the correspondence is displayed on the display, and the conditions of the sheet forming process are determined based on the display. Since the suitability can be determined, it is possible to correct the variation in the processing conditions of the sheet-shaped material forming process, and to obtain a sheet-shaped material having uniform properties in the width direction by performing uniform processing. It becomes possible.

[Brief description of the drawings]

FIG. 1 is a schematic structural view of a film manufacturing process for carrying out a method according to one embodiment of the present invention, FIG. 2 is a partial longitudinal sectional view of the base of FIG. 1, and FIG. 3 is a base of FIG. FIG. 4 is a conceptual diagram of an ON rate of a heater provided on an adjustment bolt of FIG. 4, FIG. 4 is a block diagram showing a calculation process of successive estimation using a Kalman filter, and FIG. 5 is a diagram of sheet thickness adjustment in the apparatus of FIG. FIG. 6 is a characteristic diagram of a change in thickness of a sheet-like material, showing a state; FIG. 6 is a thickness distribution diagram showing an example of a target set value of the thickness distribution of the sheet-like material; FIG. 7 is monitoring according to an embodiment of the present invention; FIG. 8 is a schematic configuration diagram of a film manufacturing process for carrying out the method, FIG. 8 is an explanatory diagram showing an example of a correspondence matrix, FIG. 9 is a graph showing contents displayed on the display means of FIG. 7, FIG. Shows the correspondence matrix for each thickness gauge position in FIG. FIG. 11 is a schematic configuration diagram of a film manufacturing process in which an arithmetic unit incorporating a Kalman filter according to the present invention and a display means are connected to a conventional thickness control system, and FIG. FIG. 13 is a schematic configuration diagram of a film manufacturing process in which an arithmetic unit and a display unit according to the present invention are applied to a system to be performed, and FIG. 13 shows a correspondence relationship between a conventional adjustment bolt of a die and a width direction position of a formed sheet. FIG. 14 is a schematic configuration diagram of a sheet-like material manufacturing process, FIG. 14 is a conceptual diagram showing an influence on a sheet-like material thickness profile when the adjustment bolt is largely changed in the apparatus of FIG. 13, and FIG. 15 is a diagram of FIG. FIG. 4 is a conceptual diagram showing a thickness change pattern of a sheet-like object when a certain adjustment bolt is operated in a minute unit in the apparatus. 21: Base 22: Melt 23: Cooling drum 24: Longitudinal stretching process 25: Width stretching process 26: Film as a sheet 27: Winders 28, 29, 30 ... Thickness gauge 31 CPU as a control device 32 Adjustment bolts as adjusting means 33 Slit-shaped discharge port 35 Heaters 53, 61, 72 CPU 50, 51, 52, 62 as an arithmetic device , 73 ... Display means

Claims (5)

(57) [Claims] 1. A melt discharged in a sheet form from a die is continuously passed through a sheet forming process, and the thickness of the sheet after passing the sheet forming process is measured in the width direction of the sheet. Measurement is performed as a distribution, and a plurality of discharge melt thickness adjusting means disposed in the width direction of the sheet of the die are controlled via a control device so that the measured value approaches a preset target value. In the method for controlling the thickness of a sheet-like material, a Kalman filter is incorporated in the control device, and the position of the die of the discharge melt thickness adjusting means in the sheet-width direction and the adjustment amount thereof are determined by using the Kalman filter. In the width direction of the sheet material after passing the material forming process, the correspondence relationship between the position where the thickness of the sheet material changes due to the adjustment of the discharge melt thickness adjusting means and the amount of change is shown. As a relational expression including a trick, an estimation is performed while sequentially correcting the correspondence matrix so that the actual thickness distribution measurement value in the sheet-like object width direction that is sequentially input approaches the target value. An optimum output to the discharge melt thickness adjusting means is determined based on the obtained relational expression including the correspondence relationship matrix, and an adjustment margin of each discharge melt thickness adjusting means is controlled based on the optimum output. A method for controlling the thickness of a sheet-shaped material. 2. A molten material discharged in a sheet form from a die is continuously passed through a sheet-like material forming process, and the thickness of the sheet-like material after passing through the sheet-like material forming process is measured in the width direction of the sheet-like material. Measured as a distribution, a sheet for controlling the thickness control means for a plurality of discharge melts disposed in the sheet width direction of the die via a control device, so that the measured value approaches a preset target value. (1) For each of the adjusting means, (i) (a) the position of the adjusting means in the discharge melt width direction and the adjustment amount of the adjusting means at a certain time; A) the position of the sheet in which the thickness changes due to the adjustment of the adjusting means in the sheet width direction after the sheet is formed by the sheet forming process, and the thickness change at the time following the time; Quantity, correspondence with Estimates the engagement determines a correspondence matrix (A * _^i-1) with respect to the respective adjusting means by its estimate, the width direction thickness distribution of the sheet in (ii) a certain time _{(t i-1) (P} i _-1) is measured, (2) the width direction thickness distribution of the sheet in the next time (t _i) of (P _i) is measured, (3) the estimated relationship matrix (a ^* _{i- 1} ) and an equation containing the measured thickness distribution (P _i-1 ), using the following formula: ( _i ) The estimated thickness distribution (P ^* _i ) at the time (t _i )
And calculating the difference between the estimated thickness distribution (P ^* _i ) and the measured thickness distribution (P _i ). (Ii) If there is the difference, the estimated correspondence matrix ( A ^* _i-1 ) is corrected to obtain a corrected estimated correspondence matrix (A ^* _i ), and the above steps (i) and (ii) are performed at each time (t _{i + 1} , t _{i + 2,} etc.). ), The estimated correspondence matrix (A ^* _i , A ^* _{i + 1,} etc.) and the measured thickness distribution ( _Pi , _{Pi + 1)}
), And repeatedly execute the Kalman filter to obtain the measured thickness distribution (P _{i + 1} , P
_{i + 2,} etc.) approximated to the target thickness distribution (). The following modified estimated correspondence matrix (A ^* _{i + 1} , A ^* _{i + 2)}
And (4) calculating the optimum adjustment amount of each adjusting means according to the target thickness distribution () from the estimated correspondence matrix obtained in step (3). Automatically controlling an adjustment amount of said adjusting means so as to obtain an optimum thickness distribution of the sheet material based on the amount. 3. A thickness distribution in a width direction of a sheet-like material formed by passing a molten material discharged from a die into a sheet shape through a sheet-like material forming process, and discharged into the die from the die in a sheet shape. A plurality of adjusting means for adjusting the thickness of the molten melt, which are arranged in the width direction of the molten melt, the method comprising: (1) for each adjusting means: The position of the adjusting means in the width direction of the melt and the adjustment amount of the adjusting means at a certain time and (b) the adjustment of the adjusting means in the sheet width direction after the sheet is formed by the sheet forming process. Estimate the correspondence between the position of the sheet-like material whose thickness changes, and the thickness change amount at the time following the above time, and estimate the correspondence matrix (A ^* _i-1 ) for all adjustment means. Ask for Was measured (ii) a certain time width direction thickness distribution of the sheet in _{(t i-1) (P} i-1), the width direction thickness of the sheet in (2) the next time (t _i) is measured distribution (P _i), using an expression that includes (3) the estimated relationship matrix (a * _^i-1) and the measured thickness distribution _{(P i-1), (} i) estimated thickness distribution at the time ^{_{(t i) (P * i}} )
And calculating the difference between the estimated thickness distribution (P ^* _i ) and the measured thickness distribution (P _i ). (Ii) If there is the difference, the estimated correspondence matrix ( A ^* _i-1 ) is corrected to obtain a corrected estimated correspondence matrix (A ^* _i ), and the above steps (i) and (ii) are performed at each time (t _{i + 1} , t _{i + 2,} etc.). ), The estimated correspondence matrix (A ^* _i , A ^* _{i + 1,} etc.) and the measured thickness distribution ( _Pi , P ^*)
_{i + 1} etc.) to repeatedly execute a Kalman filter to obtain a modified estimated correspondence matrix (A ^* _{i + 1} , A ^* _{i + 2,} etc.) at the time following them, and (4) Modified estimated correspondence matrix (A ^*
_{i + 1} , A ^* _{i + 2,} etc.) and at least the position of the adjusting means in the discharge melt width direction and the width of the sheet after passing through the sheet forming process in which the thickness changes by adjusting the adjusting means. A method for monitoring the correspondence between the thickness distribution of the sheet-like material and the thickness adjusting means, comprising displaying on a display means a correspondence relation with the position in the direction. 4. A die for discharging a molten material into a sheet, a molding process device for molding the sheet discharged from the die, and a thickness of the sheet formed by the molding process device is set to a sheet shape. Sheet thickness control comprising sheet thickness measurement means for measuring as a distribution in the product width direction, and discharge melt thickness adjustment means provided in a large number in the width direction of the die of the die. An apparatus, wherein the control device comprises:
Having a Kalman filter, using the Kalman filter, the position of the discharge melt thickness adjustment means in the width direction of the sheet of the die and the adjustment amount thereof in the width direction of the sheet after passing through the molding process device, The relationship between the position at which the thickness of the sheet-like material changes due to the adjustment of the discharge melt thickness adjusting means and the amount of the change, as a relational expression including a correspondence matrix, is sequentially input in the sheet width direction. Estimating while sequentially correcting the correspondence matrix so that the actual thickness distribution measured value approaches a preset target value, based on a relational expression including the correspondence matrix obtained by the successive estimation, the ejection is performed. The optimum output to the melt thickness adjusting means is determined, and the adjustment margin of each discharge melt thickness adjusting means is controlled based on the optimum output. Thickness controller of the sheet, characterized. 5. A sheet material is discharged from a die and continuously passed through a sheet material forming process to form a sheet material, and the thickness of the sheet material after passing the sheet material forming process is determined. Measured as a distribution in the width direction of the sheet material, and the thickness of the discharge melt provided in a large number in the width direction of the sheet material of the die via a control device so that the measured value approaches a preset target value. In a method for manufacturing a sheet-like material for manufacturing a sheet-like material while controlling an adjusting means, a Kalman filter is incorporated in the control device, and the discharge-melt thickness adjusting means is used in a sheet-like material width direction of a die using the Kalman filter. In the width direction of the sheet-like material after passing through the sheet-like material forming process, and the position where the thickness of the sheet-like material changes due to the adjustment of the discharge melt thickness adjusting means. The corresponding relationship with the change amount, as a relational expression including a corresponding relationship matrix, the corresponding relationship matrix so that the actual thickness distribution measurement value in the sheet-like object width direction sequentially inputted approaches the target value. The optimum output to the discharge melt thickness adjusting means is determined based on a relational expression including the correspondence matrix obtained by the successive estimation, and each discharge melt thickness is determined based on the optimum output. A sheet-like article is manufactured while controlling a margin of adjustment of the adjusting means.