JP2018197698A - Material estimation device, and material estimation method - Google Patents

Abstract

To provide a material estimation device and material estimation method that can estimate a material of a metal plate with high accuracy in comparison with conventional device and method.SOLUTION: A material estimation device is configured to: receive an operation condition in a processing step of processing a metal plate; give the received operation condition to an estimation model; acquire a primary estimation material serving as an estimation value of the material of the metal plate; correct the primary estimation material acquired from the estimation model on the basis of a past operation condition serving as the operation condition about the metal plate processed in the past, a past primary estimation material serving as the primary estimation material when giving a past operation condition to the estimation model, and a performance value of the material of the metal plate processed by the past operation condition; calculate a post-correction secondary order estimation material; and output a calculated secondary order estimation material.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a material estimation device and a material estimation method for estimating the material of a metal plate.

  Metal plates such as steel plates and aluminum alloy plates (Al alloy plates) are manufactured through various steps such as a melting step (steel making step if steel plate), a hot rolling step, a cooling step, a cold rolling step, and an annealing step. The The material such as hardness of the metal plate is an important quality index of the product, and the material of the product depends on the operating conditions in each of the above steps. For this reason, estimating the material of a metal plate conventionally using the operation conditions in a manufacturing process is performed. For example, in Patent Literature 1, the metal structure obtained by calculating the metal structure (precipitate amount, crystal grain size, solid solution amount of alloy element, residual stress, etc.) for each process in the production of an Al alloy plate based on the operating conditions. A method for estimating the material of an Al alloy plate based on the structure is disclosed.

JP 2002-224721 A

  However, the relationship between the operating conditions and the materials is very complicated, and it is not easy to estimate the materials from the operating conditions according to the actual manufacturing equipment. For example, in the method disclosed in Patent Document 1 described above, it is difficult to accurately calculate the metal structure in each process, and thus it is difficult to estimate the material with high accuracy.

  This invention is made | formed in view of such a situation, The main objective is to provide the material estimation apparatus and material estimation method which can solve the said subject.

  In order to solve the above-described problems, the material estimation apparatus according to one aspect of the present invention provides a material for a metal plate to be processed by the processing step when an operation condition is given in the processing step for processing the metal plate. An estimation model for estimating and outputting a primary estimation material, a past operation condition that is an operation condition for a previously processed metal plate, and a primary estimation material when the past operation condition is given to the estimation model A past record database that stores the past primary estimated material and a past record value of the material of the metal plate processed according to the previous operation condition, and a reception unit that receives the operation condition of the metal plate to be estimated. , Based on the past operation conditions corresponding to each other stored in the record database, the first primary estimated material, and the record value of the material. Further, the first estimated material output when the operating condition is given to the estimated model is corrected, the corrected second estimated material is calculated, and the first calculated by the correcting unit. An output unit that outputs the secondary estimated material.

  In this aspect, the material estimation device uses a correction function for calculating a correction value for correcting the first estimated material in accordance with the operation condition, the past operation condition corresponding to the past operation condition, and the past first order. Further comprising derivation means for deriving based on the estimated material and the actual value of the material, the correction means gives the operation condition accepted by the acceptance means to the correction function derived by the derivation means, and A correction value may be calculated, and the calculated second correction material may be calculated by applying the calculated correction value to the first estimated material.

  In the above aspect, the derivation means is configured to derive the correction function approximating a relationship between a plurality of the past operation conditions and a plurality of the past first estimated materials and the difference between the actual values. It may be.

  In the above aspect, the material estimation apparatus further includes a specifying unit that specifies the past operation condition similar to the operation condition received by the receiving unit from the result database, and the correction unit includes the specifying unit. The second estimated material may be calculated based on the past operation condition specified by the means, the past first estimated material corresponding to the past operation condition, and the actual value of the material.

  In the above aspect, the output unit outputs the first estimated material instead of the second estimated material when the past operation condition similar to the operation condition is not identified by the identifying unit. It may be configured to.

  Moreover, in the said aspect, when the said process is given with respect to the said metal plate conveyed in the longitudinal direction within the processing facility shorter than the length of the said metal plate, the said reception means is in the said metal plate. Each time a plurality of management points set along the longitudinal direction pass through a measuring device installed in the processing facility, each physical quantity at the management point measured by the measuring device is received as the operation condition. The estimation model is configured to output the first estimated material for each management point based on the operation condition at each of the management points, and the correction unit includes: Correcting each of the first estimated materials output from the estimation model for each management point, and Wherein being configured to calculate the second-order estimated material, the output unit may be configured to output the second-order estimated material for each of the control point.

  The material estimation method according to another aspect of the present invention includes a step of receiving an operation condition in a processing step of processing a metal plate, and an estimation of the material of the metal plate processed by the processing step based on the received operation condition. And providing the estimated model for outputting the first estimated material, obtaining the first estimated material, the past operating condition that is the operating condition for the metal plate processed in the past, and adding the past to the estimated model. The first estimated material that is the first estimated material when the operation condition is given, and the actual value of the material of the metal plate processed according to the past operation condition, the first obtained from the estimated model The method includes a step of correcting the primary estimated material, calculating a corrected second estimated material, and outputting the calculated second estimated material.

  According to the material estimation apparatus and the material estimation method according to the present invention, the material of the metal plate can be estimated with high accuracy based on the operation conditions.

The schematic diagram which shows the structure of the material estimation apparatus which concerns on embodiment. The block diagram which shows the structure of the material estimation apparatus which concerns on embodiment. The schematic diagram which shows the structure of a results database. The flowchart which shows the procedure of the material estimation process by the material estimation apparatus which concerns on embodiment. The schematic diagram for demonstrating classification | category of the data set of the operation condition for every management point. The flowchart which shows the procedure of similar performance data specific processing. A graph for explaining derivation of a correction function. The graph which shows the comparison result of the 1st estimated material in an evaluation test, and a track record value. The graph which shows the comparison result of the 2nd estimated material in an evaluation test, and a track record value.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In addition, each embodiment shown below illustrates the method and apparatus for actualizing the technical idea of this invention, Comprising: The technical idea of this invention is not necessarily limited to the following. Absent. The technical idea of the present invention can be variously modified within the technical scope described in the claims.

  In the present embodiment, the material estimation device uses the estimation model from the operating conditions (temperature, conveyance speed, pressure, environmental atmosphere, components, etc.) in the manufacturing process for the target metal plate, and the material (hardness, Estimate the tensile strength, elongation, etc.), correct the estimated material based on the past operating conditions, the actual material value, and the estimated material value stored in the actual database, and output the corrected material To do.

<Configuration of material estimation device>
FIG. 1 is a schematic diagram showing the configuration of the material estimation apparatus according to the present embodiment. The material estimation device 10 is for estimating the material of the metal plate 20. The metal plate 20 is long in one direction, and management points are set at a plurality of locations in the longitudinal direction. The processing facility 30 according to the present embodiment is a metal such as a rolling process (hot rolling process, cold rolling process) or a heat treatment process (annealing process, quenching process, cooling process, etc.) for the metal plate 20. This is equipment for carrying out a processing step involving conveyance of the plate 20, and its length is shorter than the length of the metal plate 20. The metal plate 20 is transported in the processing facility 30 in the longitudinal direction, and a processing step by the processing facility 30 is performed while changing a processing target portion in the longitudinal direction.

  The processing facility 30 is provided with a control device 31 that controls processing steps in the processing facility, a plurality of measuring devices 32, and a reference position detector 33. The control device 31 stores the product specifications (plate thickness, plate width, product type, chemical components, etc. when the processing process is a melting process) of the metal plate 20, and processing equipment so as to carry out the processing process according to the product specifications. 30 is controlled. The control device 31 is connected to the material estimation device 10 and outputs (transmits) product specification information to the material estimation device 10.

  The reference position detector 33 is installed in the vicinity of the entrance of the metal plate 20 of the processing facility 30 and can detect, as a reference position, the downstream end (hereinafter referred to as “leading position”) of the metal plate 20 passing immediately below in the transport direction. It is. As the reference position detector 33, for example, an optical sensor that detects that light is blocked by the passage of the metal plate 20 can be used. Moreover, when the metal plates 20 are conveyed in a state where they are connected by welding, the weld line can be detected as a reference position. In this case, as the reference position detector 33, a displacement meter or an image sensor can be used. The reference position detector 33 is connected to the material estimation device 10 and outputs reference position detection data to the material estimation device 10.

  Each measuring device 32 measures the temperature (plate temperature) of the metal plate 20 in the case of a heating or cooling process, and the rolling conditions such as the plate thickness and width, the rolling load, and the conveyance speed in the case of a rolling process. Each measuring device 32 is connected to the material estimation device 10 and outputs the measurement data to the material estimation device 10. Moreover, the test apparatus 34 which performs the material test of the metal plate 20 is provided in the vicinity of the processing facility 30, and the metal plate 20 after the processing step is provided to the test apparatus 34 to perform the material test. The test apparatus 34 is also connected to the material estimation apparatus 10, and the test result data is output to the material estimation apparatus 10.

  FIG. 2 is a block diagram showing a configuration of the material estimation apparatus 10 according to the present embodiment. The material estimation apparatus 10 is realized by a computer 11. The computer 11 includes a main body 110, an input unit 120, and a display unit 130. The main body 110 includes a CPU 111, a ROM 112, a RAM 113, a hard disk 114, an input / output interface 115, and a video output interface 116. The CPU 111, ROM 112, RAM 113, hard disk 114, input / output interface 115, and video output interface 116 are connected to each other by a bus.

  When the CPU 111 executes a material estimation program 150 that is a computer program for estimating the material of the metal plate 20, the computer 11 functions as the material estimation device 10.

  The ROM 112 stores computer programs executed by the CPU 111 and data used for the computer programs. The RAM 113 is used for reading the material estimation program 150 recorded on the hard disk 114. The RAM 113 is used as a work area for the CPU 111 when the CPU 111 executes a computer program.

  The hard disk 114 is installed with various computer programs to be executed by the CPU 111 such as an operating system and application programs, and data used for executing the computer programs. A material estimation program 150 is also installed in the hard disk 114.

The material estimation program 150 includes an estimation model 151 that is a mathematical model for estimating the material of the metal plate 20. As the estimation model 151, for example, a model represented by a mathematical formula described in JP-A-2002-126809 can be used. Specifically, a model shown in the following equation for estimating the yield point YP of the steel sheet after the rolling process can be given as an example.
YP = a.f1 + b.f2 + c.f3 + d.f4 + e
However, f1 is a component, f2 is a plate thickness, f3 is an arithmetic average of grain diameters Dγ of γ structures of recrystallized parts and non-recrystallized parts in a steel sheet after rolling, and f4 is after rolling. It is a function of the arithmetic mean of the residual strain εr of the recrystallized part and the non-recrystallized part in the steel sheet, and e is a constant.
Here, the first recrystallized part A1 and the first unrecrystallized part A2 are generated from the recrystallized part A of the steel sheet before rolling in a certain pass (the steel sheet is passed through a rolling mill), and the unrecrystallized part B When the second recrystallized part B1 and the second non-recrystallized part B2 are generated from the above, the recrystallized rate Xa of the original recrystallized part A and the recrystallized rate Xb of the original non-recrystallized part B are Calculated by
Xa = 1-exp [−0.693 (t / txr = 0.5)]
Here, txr = 0.5 = A · SvB · εsC · εD · exp (E / T)
Xb = 1-exp [−0.693 (t / txr = 0.5)]
Here, txr = 0.5 = A · SvB · εsC · (ε + εrp) D · exp (E / T)
Where t is the time between passes, txr = 0.5 is the time when the recrystallization rate is 50%, εs is the strain rate, ε is the equivalent strain, and εrp is the original (previous pass) unrecrystallized portion B , T is the rolling temperature, and Sv is the effective grain interfacial area of the crystal deformed by reduction, and is given by the following equation.
Sv = (24 / πDγ) [0.4914exp (ε) + 0.155exp (−ε) + 0.1433exp (−3ε)]
When Xa is not more than the critical recrystallization rate Xc, the γ grain size Dγ of the first recrystallized portion A1 and the residual strain εr of the first unrecrystallized portion A2 are given by the following equations.
Dγ = F · (0.2 + 0.8Xa) / (Sv ^G · ε ^H )
εr = λ · ε
Where λ = exp [−I · exp (−J / T) · t ^K · ε ^L ]
When Xa exceeds Xc, the γ grain size Dg of the first recrystallization part A1 is given by the following equation.
Dg ² = Dγ ² + M · t · exp (N / T)
However, F to N in the above equation are values determined by a separate rolling simulation experiment in accordance with the component system.
When the above Xb is equal to or less than the critical recrystallization rate Xc, the γ grain size Dγ of the second recrystallized portion B1 and the residual strain εr of the second non-recrystallized portion B2 are given by the following equations.
Dγ = F · (0.2 + 0.8Xa) / (Sv ^G · (ε + εrp) ^H )
εr = λ · (ε + εrp)
When Xb exceeds Xc, the γ grain size Dg of the second recrystallization part B1 is obtained in the same manner as A1.

  The hard disk 114 is provided with a performance database (DB) 152. FIG. 3 is a schematic diagram illustrating the structure of the performance DB 152. The record DB 152 stores record data about the metal plate 20 that has been processed by the processing facility 30 in the past. More specifically, the record DB 152 stores a record for each management point of the metal plate 20. In the result DB 152, the product ID of the metal plate 20, the management point ID, each operation condition (operation condition 1, operation condition 2,... Condition M), the results of material estimation performed in the past on the management point of the metal plate 20 (past estimation value), and the material test result (result value of material) at the management point of the metal plate 20 Is provided. The operating conditions are product specifications such as plate thickness, plate width, product type, chemical composition, and measurement values measured by the measuring instrument 32, that is, plate temperature, plate thickness, plate width, rolling load, conveyance speed, and the like. The past estimated value is a result of material estimation using an estimation model described later, and is an estimated value (first estimated material) before correction.

  Refer to FIG. 2 again. An input unit 120 including a keyboard and a mouse is connected to the input / output interface 115. Further, the control device 31, the measuring device 32, the reference position detector 33, and the test device 34 are connected to the input / output interface 115. Data output from the control device 31, the measuring device 32, the reference position detector 33, and the test device 34 is given to the CPU 111 via the input / output interface 115.

  The video output interface 116 is connected to a display unit 130 configured by an LCD, a CRT, or the like, and outputs a video signal corresponding to video data given from the CPU 111 to the display unit 130. The display unit 130 displays an image according to the input video signal.

<Operation of material estimation device>
Next, the operation of the material estimation apparatus 10 will be described. FIG. 4 is a flowchart showing a procedure of material estimation processing by the material estimation apparatus 10 according to the present embodiment. In the execution of the processing step, the control device 31 controls the processing facility 30. Thereby, the metal plate before a process is conveyed, it approachs into the processing equipment 30 from the head, and a processing process is implemented in the processing equipment 30. The control device 31 outputs the product specifications to the material estimation device 10, the reference position detector 33 outputs the detection data of the reference position of the metal plate 20 to the material estimation device 10, and each measuring device 32 includes the plate temperature and the plate thickness. The sheet width, rolling load, conveyance speed, etc. are measured, and the measurement result is output to the material estimation apparatus 10.

したがって、ＣＰＵ１１１は、基準位置検出器３３が基準位置２１の検出データを受信した時刻ａから、式（１）にしたがって時刻ａ＋ｋを算出し、この時刻ａ＋ｋにおける測定器３２の測定結果を当該管理ポイント２２の操業条件のデータセットに含める。ＣＰＵ１１１は、このような処理を全管理ポイント２２及び全測定器３２について実行することで、全ての管理ポイント２２についての操業条件のデータセットを作成する。但し、製品仕様のような金属板２０で１つ与えられる操業条件は、全ての管理ポイント２２のデータセットに含められる。 The CPU 111 of the material estimation device 10 receives the product specifications output from the control device 31 and the measurement data output from each measuring device 32 as operating conditions (step S101). Next, the CPU 111 classifies the accepted operation conditions into a data set for each management point (step S102). Here, the process of step S102 will be described. FIG. 5 is a schematic diagram for explaining the classification of the operation condition data set for each management point. In the processing facility 30, the metal plate 20 is transported in the transport direction at a plate passing speed S (t). A reference position detector 33 is installed at a distance D away from the upstream side of the measuring device 32 in the transport direction. In addition, it is assumed that the management point 22 is set at a position separated by a distance L upstream of the reference position 21 in the transport direction. When the time when the reference position 21 of the metal plate 20 passes directly under the reference position detector 33 is a and the time when the management point 22 passes directly under the measuring device 32 is a + k, the following equation (1) is established.

Therefore, the CPU 111 calculates the time a + k according to the equation (1) from the time a when the reference position detector 33 receives the detection data of the reference position 21, and the measurement result of the measuring device 32 at the time a + k is calculated as the management point. It is included in the data set of 22 operating conditions. The CPU 111 executes such processing for all the management points 22 and all the measuring instruments 32, thereby creating a data set of operation conditions for all the management points 22. However, one operation condition given by the metal plate 20 such as a product specification is included in the data set of all the management points 22.

  Next, the CPU 111 selects one management point 22 as a material estimation target (step S103). Hereinafter, the selected management point 22 is referred to as an “estimation target management point”. Further, the CPU 111 applies the operation condition data set of the estimation target management point 22 to the estimation model 151, and calculates a first estimated material that is an estimated value of the material (step S104).

但し、ｄ_ｉは推定対象データと実績データｉの過去操業条件との距離を示す。ここでの距離には、ユークリッド距離、マハラノビス距離等を使用できる。また、ｊは操業条件の番号を示し、ｔｇｔは推定対象の金属板２０を示し、Ｘ_ｉｊは実績データｉにおける過去操業条件ｊを示し、Ｘ_{ｔｇｔ＿ｊ}は推定対象データにおける操業条件ｊを示し、ａ_ｊは距離を求めるためのパラメータであり、Ｃは類似度を求めるためのパラメータである。パラメータａ_ｊ,Ｃは物理的知見又は統計解析結果に基づいて定めてもよいし、後述するステップＳ１０９において算出される第２次推定材質の二乗誤差が最小になるように遺伝的アルゴリズム、ＰＳＯ（Particle Swarm Optimization）等の最適化計算により定めてもよい。 Next, the CPU 111 executes a similar result data specifying process for specifying result data including an operation condition similar to the input operation condition for the estimation target management point (step S105). The similar result data specifying process will be described in detail. FIG. 6 is a flowchart showing the procedure of the similar result data specifying process. In the similar result data specifying process, first, the CPU 111 sets a variable i indicating a record number of the result DB 152 to “1” (step S1051), and an operation condition at the estimation target management point (hereinafter referred to as “estimation target data”). The degree of similarity between the i-th record (hereinafter referred to as “actual data i”) in the result DB 152 and the same operation condition as described above is calculated (step S1052). Similarity _{w i} is calculated using the following equation (2) and (3).

However, d _i indicates the distance between the estimation target data and the past operation condition of the performance data i. As the distance here, Euclidean distance, Mahalanobis distance, or the like can be used. Further, j indicates the number of the operation condition, tgt indicates the metal plate 20 to be estimated, X _ij indicates the past operation condition j in the performance data i, X _{tgt_j} indicates the operation condition j in the estimation target data, a _j is a parameter for _obtaining the distance, and C is a parameter for obtaining the similarity. The parameters a _j , C may be determined based on physical knowledge or statistical analysis results, or a genetic algorithm, PSO (PSO ( It may be determined by optimization calculation such as Particle Swarm Optimization.

Then CPU111 compares the similarity _{w i} and a predetermined reference value, determines whether the degree of similarity _{w i} is equal to or greater than the reference value (step S 1053). This reference value may be determined according to the experience of the operator, or may be determined by optimization calculation such as a genetic algorithm or PSO so that the square error of the second estimated material is minimized. If similarity score w _i is greater than or equal to the reference value (YES in step S1053), performance data i is specified as performance data similar to estimation target data (hereinafter referred to as “similar performance data”), and stored in hard disk 114. (Step S1054), the process proceeds to step S1055. On the other hand, if similarity score w _i is less than the reference value (NO in step S1053), CPU 111 moves the process directly to step S1055.

The CPU 111 determines whether i has reached the maximum value (the number of products registered in the result DB 152) P (step S1055). If not (NO in step S1055), i is incremented. (Step S1056), the process returns to step S1052. Thus, CPU 111 while changing the i from 1 to P, and calculates the similarity w _i, similarity w _i to identify the similar actual data equal to or larger than the reference value. If i has reached P (YES in step S1055), CPU 111 ends the similar result data specifying process.

但し、Ｍは操業条件数を示し、β_０は定数であり、β_１,…,β_Ｍは計数である。ここで、過去推定値と実績値との差をｙとおき、次式（５）のｅｒｒが最小になるβ_１,…,β_Ｍを算出することで、補正関数を導出する。

但し、Ｎは抽出された類似実績データ数を示す。 Refer to FIG. 4 again. After the similar record data specifying process, the CPU 111 determines whether or not one or more similar record data is specified (step S106). When one or more similar performance data is specified (YES in step S106), the CPU 111 extracts similar performance data from the performance DB 152, and describes the operation condition (past operation conditions) of the extracted similar performance data as an explanatory variable, A correction function is derived using the similarity as a weight and the difference between the past estimated value and the actual value as an objective variable (step S107). More specifically, weighted multiple regression is used to derive this correction function. That is, the correction function f _H is defined as the following equation (4).

Here, M represents the number of operating conditions, β ₀ is a constant, and β ₁ ,..., Β _M is a count. Here, the difference between the past estimated value and the actual value is set as y, and β ₁ ,..., Β _M that minimizes err in the following equation (5) are calculated to derive a correction function.

However, N shows the number of similar performance data extracted.

FIG. 7 is a graph for explaining the derivation of the correction function. In FIG. 7, the vertical axis represents the objective variable, and the horizontal axis represents the explanatory variable. Also, white circles indicate performance data that is not similar performance data, and black circles indicate similar performance data. The calculation of β ₁ ,..., Β _M described above is performed by using an approximate function that approximates the relationship between the past operation condition in the similar performance data and the difference y between the past estimated value and the actual value by the least square method. Means a straight line shown).

Refer to FIG. 4 again. When the correction function is derived, the CPU 111 calculates the correction value by inputting the estimation target data to the correction function (step S108), and the second-order estimation that is the corrected material estimated value from the following equation (6). The material is calculated (step S109).
Secondary estimated material = primary estimated material-correction value (6)

  Next, the CPU 111 displays the calculated second-order estimated material on the display unit 130 (step S110). As a result, the second estimated material after correction based on the similar result data is presented to the operator. After outputting the second estimated material, the CPU 111 moves the process to step S112.

  If no similar record data is specified in step S106 (NO in step S106), the CPU 111 calculates the first estimated material calculated in step S104 without calculating the second estimated material. It is displayed on the display unit 130 (step S111). When there is no similar result data, if a correction function based on the result data is derived to correct the first estimated material, the estimation error may be increased. Therefore, when past operation conditions similar to the current operation condition are not registered in the performance DB 152, the first estimated material that has not been corrected is output. As a result, an estimated value with a minimum estimation error can be presented to the operator. After outputting the first estimated material, the CPU 111 moves the process to step S112. The operator confirms the output first or second estimated material, reviews the operating conditions of the processing step, operates the control device 31 to change the operating conditions as appropriate.

  The CPU 111 registers the product ID and management point ID, each operation condition, and the material estimated value of the metal plate 20 to be estimated in the performance DB 152 (step S112). As a result, the result data is accumulated in the result DB 152. Note that the estimated material value registered in the performance DB 152 is the second estimated material when the second estimated material is calculated, and the first estimated material when the second estimated material is not calculated. Estimated material.

  Next, the CPU 111 determines whether or not all management points 22 have been selected as estimation target management points (step S113), and if there are any management points 22 that have not yet been selected (NO in step S113). Then, the processing is returned to step S101, the newly input operation condition is accepted (step S101), the operation condition is classified into a data set for each management point (step S102), and one of the management points 22 that has not been selected yet. Is selected as an estimation target management point (step S103), and the processing from step S104 onward is executed. When selection of all management points 22 on metal plate 20 is completed (YES in step S113), CPU 111 moves the process to step S114.

  When the end of the metal plate 20 passes through the processing equipment 30 and the processing by the processing process is completed over the entire length of the metal plate 20, a material test is performed on each management point 22 of the metal plate 20 by the test device 34. When the material test is performed, the test device 34 transmits the test result to the material estimation device 10 together with the product ID and the management point ID. The CPU 111 receives the test result transmitted from the test apparatus 34 (step S114), and registers the test result in the result DB 152 in association with the product ID and the management point ID (step S115). This completes the material estimation process.

  By configuring as described above, the material estimation apparatus according to the present embodiment approximates the relationship between the operating condition and the difference between the past estimated value and the actual value of the first estimated material calculated by the estimation model. Since the correction is performed using the corrected function, the material can be estimated with high accuracy.

  Moreover, the management point 22 is set in the longitudinal direction of the metal plate 20, the operation condition is acquired for each management point 22, and the second estimated material is calculated, so that the operation is performed over the entire length of the very long metal plate 20. Even when the conditions are not uniform and the material after treatment changes depending on the location in the longitudinal direction of the metal plate 20, the material at each management point 22 can be estimated.

<Evaluation test>
The inventors conducted a performance evaluation test of the material estimation method according to the present embodiment. In this test, the metal plate 20 is a steel plate, the treatment process is an annealing process, the estimated material is the tensile strength, the first estimated material and the second estimated material are calculated, and the first estimated material and the actual value are calculated. Comparison and comparison between the second estimated material and the actual value were performed. FIG. 8A is a graph showing a comparison result between the first estimated material and the actual value in this test, and FIG. 8B is a graph showing a comparison result between the second estimated material and the actual value. In FIG. 8A, the vertical axis represents the primary estimated material, the horizontal axis represents the actual value, and the straight line drawn diagonally is a straight line where “first estimated value = actual value”. In FIG. 8B, the vertical axis indicates the secondary estimated material, the horizontal axis indicates the actual value, and the straight line drawn diagonally is a straight line where “secondary estimated value = actual value”. As shown in FIGS. 8A and 8B, the second estimated material has a smaller variation in predicted values than the first estimated material, and is closer to the actual value. Therefore, it can be seen that the material can be estimated with high accuracy by correcting the primary estimated material.

(Other embodiments)
In the above-described embodiment, the calculated second estimated material is displayed on the display unit 130. However, the output form of the second estimated material is not limited to this. It can also be set as the structure which prints a 2nd estimated material with a printing apparatus. Further, the second estimated material itself can be output to another device instead of being presented to the operator, or can be output from the material estimation program 150 to another computer program. For example, the CPU 111 executes a computer program for calculating a change value of the operation condition for bringing the estimated value of the material close to the material of the product specification, and outputs the calculated second estimated material to the computer program. It can also be configured to calculate a change value of the condition. In this case, by outputting a change value of the operation condition and notifying the operator, the operator can change the operation condition according to the change value. In addition, the material estimation device 10 outputs (transmits) the second estimated material to the control device 31, so that the control device 31 calculates the control amount of the operation condition according to the second estimated material, and automatically controls the processing process. It can also be set as the structure which performs.

  In the above-described embodiment, the configuration in which all processing of the material estimation program 150 is executed by the single computer 11 is described. However, the present invention is not limited to this, and the material estimation program 150 and It is also possible to provide a distributed system in which similar processing is executed by a plurality of devices (computers).

  The material estimation apparatus and material estimation method of the present invention are useful as a material estimation apparatus and material estimation method for estimating the material of a metal plate.

DESCRIPTION OF SYMBOLS 10 Material estimation apparatus 11 Computer 120 Input part 130 Display part 111 CPU
112 ROM
113 RAM
114 hard disk 150 material estimation program 151 estimation model 152 results database 20 metal plate 22 management point 30 processing facility 32 measuring instrument 34 test apparatus

Claims (7)

An estimation model for estimating a material of the metal plate to be processed by the processing step and outputting a first estimated material when operating conditions in the processing step of processing the metal plate are given;
According to past operating conditions that are operating conditions for metal plates that have been processed in the past, past first estimated materials that are primary estimated materials when the past operating conditions are given to the estimated model, and the past operating conditions A performance database that stores the performance values of the processed metal plate material in association with each other;
Accepting means for accepting the operating conditions related to the metal plate of the material estimation target;
Based on the past operation conditions corresponding to each other stored in the record database, the first primary estimated material, and the record value of the material, the operation condition received by the receiving unit is given to the estimation model. Correction means for correcting the first estimated material that is sometimes output and calculating a corrected second estimated material;
An output unit that outputs the second estimated material calculated by the correction unit;
Material estimation device. A correction function for calculating a correction value for correcting the first estimated material according to the operation condition is changed to the corresponding past operation condition, the past first estimated material, and the actual value of the material. Derivation means for deriving based on
The correction means calculates the correction value by giving the operation condition received by the reception means to the correction function derived by the derivation means, and uses the calculated correction value as the first estimated material. Configured to apply and calculate the second estimated material;
The material estimation apparatus according to claim 1. The derivation means is configured to derive the correction function approximating a relationship between a plurality of the past operation conditions and a plurality of the past first estimated materials and the difference between the actual values,
The material estimation apparatus according to claim 2. From the results database, further comprising a specifying means for specifying the past operating conditions similar to the operating conditions received by the receiving means,
The correction means calculates the second estimated material based on the past operation condition specified by the specifying means, the corresponding past first estimated material and the actual value of the material corresponding thereto. It is configured,
The material estimation apparatus in any one of Claims 1 thru | or 3. The output unit is configured to output the first estimated material instead of the second estimated material when the past operation condition similar to the operation condition is not identified by the identifying unit. Yes,
The material estimation apparatus according to claim 4. The receiving means is arranged along the longitudinal direction of the metal plate when the processing step is performed on the metal plate conveyed in the longitudinal direction in a processing facility shorter than the length of the metal plate. Each time a plurality of set management points pass through a measuring device installed in the processing facility, each physical quantity at the management point measured by the measuring device is received as the operation condition. ,
The estimation model is configured to output the first estimated material for each management point based on the operation conditions at each of the management points,
The correction means is configured to correct each of the first estimated materials output from the estimated model for each management point, and calculate the second estimated material at each of the management points. ,
The output unit is configured to output the second estimated material for each management point.
The material estimation apparatus in any one of Claims 1 thru | or 5. Receiving the operating conditions in the processing step of processing the metal plate;
Estimating the material of the metal plate to be processed by the processing step, giving the received operating condition to an estimation model that outputs a first estimated material, and obtaining the first estimated material;
According to past operating conditions that are operating conditions for metal plates that have been processed in the past, past first estimated materials that are primary estimated materials when the past operating conditions are given to the estimated model, and the past operating conditions Correcting the first estimated material acquired from the estimated model based on the actual value of the processed metal plate material, and calculating a corrected second estimated material;
Outputting the calculated second estimated material, and
Material estimation method.

Images