TW201235124A - Energy consumption prediction device - Google Patents

Energy consumption prediction device Download PDF

Info

Publication number
TW201235124A
TW201235124A TW100109076A TW100109076A TW201235124A TW 201235124 A TW201235124 A TW 201235124A TW 100109076 A TW100109076 A TW 100109076A TW 100109076 A TW100109076 A TW 100109076A TW 201235124 A TW201235124 A TW 201235124A
Authority
TW
Taiwan
Prior art keywords
value
energy consumption
calculation
energy
learning
Prior art date
Application number
TW100109076A
Other languages
Chinese (zh)
Other versions
TWI483790B (en
Inventor
Kazutoshi Kitagou
Original Assignee
Toshiba Mitsubishi Elec Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to JP2011033002A priority Critical patent/JP5666338B2/en
Application filed by Toshiba Mitsubishi Elec Inc filed Critical Toshiba Mitsubishi Elec Inc
Publication of TW201235124A publication Critical patent/TW201235124A/en
Application granted granted Critical
Publication of TWI483790B publication Critical patent/TWI483790B/en

Links

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B21MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21BROLLING OF METAL
    • B21B37/00Control devices or methods specially adapted for metal-rolling mills or the work produced thereby
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B21MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21BROLLING OF METAL
    • B21B38/00Methods or devices for measuring, detecting or monitoring specially adapted for metal-rolling mills, e.g. position detection, inspection of the product
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions

Abstract

An energy consumption prediction device with high prediction accuracy for a hot rolling line is provided. The energy consumption prediction device includes: an energy consumption calculation device which calculates an calculated value of energy consumption using setting values of rolling torque, roll speed, and rolling power; an energy consumption actual value calculation device which calculates a measured and calculated value of energy consumption using a calculated value calculated from measured values of rolling torque and roll speed; an energy consumption actual value acquisition device which acquires an actual value of energy consumption by integrating measured rolling powers measured in action; an energy consumption learning value calculation device which calculates a learning value of energy consumption by comparing the measured and calculated value of energy consumption and the actual value of energy consumption; and a prediction value calculation device which calculates a prediction value of energy consumption obtained by reflecting the learning value of energy consumption to the calculated value of energy consumption.

Description

201235124 六、發明說明: 【發明所屬之技術領域】 本發明係關於預測製造金屬製品之熱軋作業線中的能 源消費量之能源消費量預測裝置。 【先前技術】 利用熱軋作業線來製造希望的尺寸、品質之製品所需 消耗的能源消費量’係使用例如軋延機架(r〇Uing stand) 的軋延轉矩及軋輥速度來算出(參照例如專利文獻丨)。另 外,提出有:以針對每一被軋延材概略決定軋延轉矩及軋 軺i速度為則提,不使用軋延轉矩及軋輥速度的預測值,而 針對每一個以材質、軋延時間及軋延前後的被軋延材尺寸 為根據所做的區分來學習能源消費量之方法(參照例如專 利文獻2)。 【先前技術文獻】 (專利文獻) 專利文獻1 :日本發明專利第3444267號公報 專利文獻2:日本發明專利第3498786號公報 【發明内容】 (發明所欲解決之課題) 在能源消費量的預測所使用的參數、亦即軋延時間、 軋延轉矩、軋輥速度之中,軋延轉矩可藉由使用模型公式 (model formula)之設定計算等來精確度良好地預測。然 而,軋延時間及軋輥速度在實際的軋延中與預測值之間容 易產生誤差,而成為能源消費量的預測誤差的主要原因。 322914 4 201235124 此外’能源消費量還會因為在設定計算及設定計算學習上 並未將驅動軋延機架的軋輥之馬達的特性會隨時間經過而 變化等因素考慮進去的原因而變化。因此,為了正確地預 測能源消費量’有必要藉由學習計算來修正上述種種誤差。 然而’針對每一個以材質、軋延時間及軋延前後的被 軋延材尺寸為根據所做的區分來學習能源消費量之上述方 法’並未使用軋延轉矩及軋輥速度的預測值。所以,即使 是同一區分,也會有軋延條件變化時,軋延轉矩及軋輥速 度一旦變化’預測精確度就會降低之問題。 本發明係鑑於上述問題點而完成者,其目的在提供一 種預測精確度高之熱軋作業線的能源消費量預測裝置。 (解決課題之手段) 根據本發明的一個態樣,提供一種熱軋作業線之能源 消費量預測裝置,具備有:取得在熱軋作業線的軋延處理 中所測量的動作實測值之實測值取得裝置;比較將動作實 冽值應用至模型公式的參數而得到的動作實測計算值與動 作實測值,而算出設定計算學習值之設定計算學習裝置; =用熱軋作業料作業條件及設定計算學習值,來計算包 3熱軋作業線中的軋延轉矩、軋親速度及軋延功率的設定 的動作設定值之設定計算裝置;使用動作設定值來 H肖費量計算值之能源消費量算出裝置;使用乳延 及軋㈣度的動作實測計算值來算出能源;肖費量實測 動能源消^量實際值算出裝置;藉由將札延功率的 d值予以積分來取得能源消費量實際值之能源消費 322914 5 201235124 置;藉由比較能源消費量實測計算值及能 學、習值算出:以算,源::費量學習值之能源消費量 以及算出使費量學習值反映至能 =費董計算值而得到的能源消費量預測值之預測值算月出匕 裝置。 (發明之效果) 線的=:預=供一種預測精確度高之熱軋作業 【實施方式】 接著’參照圖式來說明本發明之第一至第四實施形 態。在以下的圖式的記載中,相同或類似的部份均標以相 同或類似的符號。以下所示的實施形態,係舉例來說明用 來使本發明的技術思想具體化之裝置或方法者,本發明之 實施形態之構成構件的構造、配置等並不限定於以下所述 者。本發明之實施形態可在申請專利範圍内施加各種變更。 (第一實施形態) 本發明第一實施形態之能源消費量預測裝置10,係預 測熱軋作業線20的能源消費量之裝置,如第i圖所示,具 備有··實測值取得裝置11、設定計算學習裝置12、設定計 算裝置13、能源消費量算出裝置14、能源消費量實際值算 出裝置15、能源消費量實際值取得裝置16、能源消費量學 習值算出裝置17、預測值算出裝置丨8。 實測值取得裝置11 ’係取得在熱軋作業線20的軋延 處理中所測量之包含軋延轉矩、軋輥速度及遍及被軋延材 6 322914 201235124 - 的全長的軋延功率等之實測值(以下,稱之為「動作實測值 、 AACT」)。 設定計算學習裝置12,係比較將動作實測值Aagt應用 至模型公式的參數而得到的動作實測計算值AAeTeAL與動作 實測值Aagt,而算出設定計算學習值ZnM。此處,上標AC:T 表示動作實測值,上標ACTCAL表示動作實測計算值(以下 的說明中皆同)。 設定計算裝置13,係使用熱軋作業線20的作業條件 及設定計算學習值ZnM,來計算包含熱軋作業線20中的軋 延轉矩、軋輥速度及軋延功率的設定值在内之動作設定值 ASET。能源消費量算出裝置14,係使用動作設定值ASET來算 出能源消費量計算值EnSET。 能源消費量實際值算出裝置15,係使用軋延轉矩及軋 輥速度的動作實測計算值AAeKU來算出能源消費量實測計 算值EnAeT·。另一方面,能源消費量實際值取得裝置16則 是將藉由實測值取得裝置11而取得之軋延功率的動作實 測值PwAeT予以積分來取得能源消費量實際值EnACT。 能源消費量學習值算出裝置17,係藉由比較能源消費 量實測計算值EnAeTm及能源消費量實際值ΕηΑσΓ,來算出能 源消費量學習值ZnECeuR。所算出的能源消費量學習值ZnECeUR 係儲存至能源消費量學習值儲存裝置171。 預測值算出裝置18,係算出使從能源消費量學習值儲 存裝置171讀出的能源消費量學習值ZnECeUR反映至能源消 費量計算值EnSET而得到的能源消費量預測值EnPl"ed。 7 322914 201235124 第2圖係顯示作為能源消費量預測裝置10進行能源消 費量預測的預測對象之熱軋作業線2G的構成例。第2圖所 不之熱札作業線20具有加熱爐2卜粗軋機23、精軋機26、 及捲取機28。 從加熱爐21搬出的被軋延材100,係由可逆式的粗軋 機23加以乾延。報軋機23通常具有一台至數台之軋延機 架(rolling stand) ’且藉由使被軋延材1〇〇往復移動並通 過粗乳機23數次’以在粗軋機的出口侧將被軋延材100軋 延到目標的中間條板厚度。以下將「使被軋延材通過 粗札機23的軋延機架」這件事稱為「道次(pass)」。 經粗軋機23加以軋延後,將被軋延材丨⑽從粗軋機 23的出口側搬送至精軋機%的入口側,利用由例如5至7 台軋延機架260構成之精軋機26將被軋延材100軋延到所 希望的製品板厚度。從精軋機26搬出之被軋延材1〇〇,經 水冷裝置等冷卻裝置27加以冷卻之後,由捲取機28加以 捲繞成線圈狀。 粗軋機23的軋延機架的軋輥係由馬達231所驅動,精 軋機26的軋延機架260的軋輥係由馬達261所驅動。另 外,在粗軋機23的入口側配置有粗軋機入口側去銹皮器 (descaler)22’在精軋機26的入口側配置有精軋機入口側 去銹皮器25。並且,在粗軋機23與精軋機26間之搬送台 區域配置有盤捲箱(coil b〇x;)24。 第1圖所示之能源消費量預測裝置1〇,係算出為了製 造所希望的尺寸、品質之製品所必需之熱軋作業線2〇的能 8 322914 201235124 " 源消費量的預測值。以下,說明能源消費量預測裴置丨 、詳細的動作。 設定計算裝置13,係根據作業條件及設定計算學 Zn«,使用公知的模型公式來計算動作設定值AsET。作 件係為例如在精軋機26的出口側所要達到之目標板厚” 軋機的出口侧溫度等。另外,動作設定值係針許為= 使被軋延材100成為所希望的板厚所必需的軋輥間隙:以 及為了實現所希望的精軋機出口測溫度所必需的軋輥速卢 等而算出。亦即,為了使在熱軋作業線2〇製造出之製品f 現所希望的尺寸及品質所必需的軋延轉矩、軋輥速度等 係由設定計算裝置13加以計算出。 再者,在不超過精軋機26的負荷極限值、及不超過驅 動軋延機架260的軋輥之馬達261的轉矩極限值之條件 下,藉由設定計算裝置13計算出包含各道次及各軋延機架 260的軋延負荷、軋延轉矩及軋輥速度等的設定值在内之 動作設定值ASET。 此專動作设疋值A ’最好在至少:從被札延材10〔丨 的前端算起之就精確度的確保而言很重要之咬入點、從確 保生產量的點來說使被軋延材100加速之軋延速度最大之 中間點、及被軋延材100的溫度變低之尾端點這三個點算 出。以下將算出動作設定值ASET的地點稱為「標的點」。 此外,將上述之一連串計算稱為「設定計算」。藉由設 定計算而計算出的軋輥間隙及軋輥速度等之動作設定值 ASET,係輸出至熱軋作業線20的控制裝置,熱軋作業線20 9 322914 201235124 根據此等動作設定值ASET而進行作業。另外,軋延轉矩、 軋輥速度之動作設定值AsET,係作為能源消費量計算用的參 數而輸出至能源消費量算出裝置14。 實測值取得裝置li,係從熱軋作業線20所設置的量 測器(省略圖示)取得遍及軋延處理中之被軋延材1 〇〇全長 的軋延轉矩、軋輥逮度、及軋延功率等之動作實測值aact。 舉例來說’軋延轉矩係使用施加在軋輥之負荷等來算出, 軋輥速度係使用軋輥的轉速等來算出,軋延功率係使用馬 達261的驅動電流等來算出。 藉由實測值取得裝置Π而取得之動作實測值AACT,係 輸出至設定計算學習裝置。 設定計算學習裝置12,係將實測值取得裝置11所取 得之動作實測值Aact代入模型公式的參數而算出動作實測 計算值Aactcal。此外,設定計算學習裝置12係藉由比較實 測值取得裝置11所取得之被軋延材1〇〇的各標的點的動作 實測值Aact、與算出的動作實測計算值Aactcal,來學習動作 實測值ΑΑα與動作實測計算值AAaeAL的誤差。 具體而言,設定計算學習裝置12係算出動作實測值 AAeT之相對於動作實測計算值AKTGAL之比。換言之,設定計 算學習裝置12係算出設定計算學習值ZnM,以作為「動作 實測值AAeT/動作實測計算值AAa(:AL j。 算出的設定計算學習值ΖηΜ係儲存至設定計算學習值 儲存裝置121。儲存至設定計算學習值儲存裝置121之設 定計算學習值ΖηΜ,係供設定計算裝置13使用。 10 322914 201235124 — 能源消費量算出裝置14所進行之熱軋作業線20的能 、 源消費量之計算中,會使用到由設定計算裝置13所算出之 動作設定值ASET。具體而言,能源消費量算出裝置14係根 據算出的軋延轉矩及軋輥速度等之動作設定值ASET、及作業 條件,來算出在粗軋機23中的各道次及精軋機26的各軋 延機架260的能源消費量計算值EnSET。舉例來說,各道次 及驅動各軋延機架260的軋輥之馬達261的能源消費量計 算值EnSET,係如以下之式(1)及式(2)所示,藉由對軋延轉 矩G(t)[kNm]及軋輥速度v(t)[m/s]之積進行時間(t)[s] 積分而算出:201235124 VI. Description of the Invention: [Technical Field of the Invention] The present invention relates to an energy consumption prediction device for predicting energy consumption in a hot rolling line for manufacturing metal products. [Prior Art] The energy consumption required to produce a desired size and quality product by using a hot rolling line is calculated using, for example, a rolling stand and a roll speed of a rolling stand ( See, for example, the patent document 丨). In addition, it is proposed that the rolling torque and the rolling speed are determined for each rolled material, and the predicted values of the rolling torque and the rolling speed are not used, and the material is rolled for each material. The time and the size of the rolled material before and after the rolling are a method of learning the energy consumption amount based on the division (see, for example, Patent Document 2). [Prior Art Document] (Patent Document) Patent Document 1: Japanese Patent No. 3444267 Patent Document 2: Japanese Patent No. 3498786 (Invention) The problem to be solved by the invention is in the prediction of energy consumption. Among the parameters used, that is, rolling time, rolling torque, and roll speed, the rolling torque can be accurately predicted by calculation using a model formula or the like. However, the rolling time and the roll speed are prone to errors between the actual rolling and the predicted value, and become the main cause of the prediction error of energy consumption. 322914 4 201235124 In addition, the energy consumption will also be changed because factors such as the characteristics of the motor that drives the rolls of the rolling stand will change over time in setting calculations and setting calculations. Therefore, in order to correctly predict energy consumption, it is necessary to correct the above errors by learning calculations. However, the above method of learning energy consumption for each material based on the material, the rolling time, and the size of the rolled material before and after rolling is not used as the predicted value of rolling torque and roll speed. Therefore, even if the same classification is made, there will be a problem that the rolling accuracy and the roll speed change once the rolling condition changes, and the prediction accuracy is lowered. The present invention has been made in view of the above problems, and an object thereof is to provide an energy consumption amount predicting apparatus for a hot rolling line having high prediction accuracy. (Means for Solving the Problem) According to an aspect of the present invention, an apparatus for predicting an energy consumption amount of a hot rolling line is provided, comprising: obtaining an actual measured value of an actual measured value measured in a rolling process of a hot rolling line Acquiring the device; comparing the calculated measured value and the measured value of the action obtained by applying the actual value of the action to the parameter of the model formula, and calculating the setting calculation learning device for setting the calculated learning value; = calculating the working condition and setting of the hot rolling material Learning value, a calculation setting device for calculating the setting value of the rolling torque, the rolling contact speed, and the rolling power in the hot rolling line of the package 3; and the energy consumption calculated by using the operation setting value The calculation device uses the measured values of the emulsion extension and the rolling (four degrees) to calculate the energy; the actual energy calculation device for calculating the dynamic energy consumption; and the energy consumption is obtained by integrating the d value of the Zayen power Actual value of energy consumption 322914 5 201235124; by comparing the measured value of energy consumption and the calculated and learned value: to calculate, source:: the amount of learning value Prediction source consumption and calculated the amount of the fee to be able to reflect the learning value = fee calculated values obtained by Dong Energy Consumption value calculation of the predicted value moonrise dagger device. (Effects of the Invention) The following is a description of the first to fourth embodiments of the present invention. In the description of the following drawings, the same or similar parts are denoted by the same or similar symbols. In the following embodiments, the apparatus or method for embodying the technical idea of the present invention is exemplified, and the structure, arrangement, and the like of the constituent members of the embodiment of the present invention are not limited to the following. The embodiments of the present invention can be modified in various ways within the scope of the patent application. (First Embodiment) The energy consumption amount prediction device 10 according to the first embodiment of the present invention is an apparatus for predicting the energy consumption amount of the hot rolling operation line 20, and as shown in Fig. i, the actual value acquisition device 11 is provided. The calculation calculation device 12, the setting calculation device 13, the energy consumption amount calculation device 14, the energy consumption amount actual value calculation device 15, the energy consumption amount actual value acquisition device 16, the energy consumption amount learned value calculation device 17, and the predicted value calculation device丨 8. The measured value obtaining device 11' obtains the measured value including the rolling torque, the roll speed, and the rolling length of the entire length of the rolled material 6 322914 201235124 - measured in the rolling process of the hot rolling line 20 (Hereinafter, it is called "action measured value, AACT"). The setting calculation learning means 12 compares the operation actual measurement calculation value AAeTeAL and the operation actual measurement value Aagt obtained by applying the motion measurement value Aagt to the parameter of the model formula, and calculates the set calculation learning value ZnM. Here, the superscript AC:T represents the measured value of the action, and the superscript ACTCAL represents the measured value of the action (the same applies in the following description). The setting calculation device 13 calculates the operation value including the rolling torque, the roll speed, and the rolling power in the hot rolling line 20 by using the working conditions of the hot rolling line 20 and the setting calculation learning value ZnM. Set the value ASET. The energy consumption amount calculation means 14 calculates the energy consumption amount calculation value EnSET using the operation setting value ASET. The energy consumption actual value calculation means 15 calculates the energy consumption measurement actual value EnAeT· using the measured value AAKU of the rolling torque and the roll speed. On the other hand, the energy consumption actual value acquisition means 16 integrates the actual operation value PwAeT of the rolling power obtained by the actual value acquisition means 11 to obtain the energy consumption actual value EnACT. The energy consumption learning value calculation means 17 calculates the energy consumption learning value ZnECeuR by comparing the energy consumption measured value EnAeTm and the energy consumption actual value ΕηΑσΓ. The calculated energy consumption learning value ZnECeUR is stored in the energy consumption learned value storage device 171. The predicted value calculation means 18 calculates the energy consumption amount predicted value EnPl " ed obtained by reflecting the energy consumption amount learned value ZnECeUR read from the energy consumption amount learned value storage means 171 to the energy consumption amount calculation value EnSET. 7 322914 201235124 Fig. 2 shows an example of the configuration of the hot rolling line 2G to be predicted by the energy consumption amount predicting device 10 for energy consumption amount prediction. In the second drawing, the hot work line 20 includes a heating furnace 2, a roughing mill 23, a finishing mill 26, and a coiler 28. The rolled material 100 taken out from the heating furnace 21 is dried by a reversible roughing mill 23. The newspaper rolling mill 23 usually has one to several rolling stands 'and will be reciprocated by the rolled material 1 and passed through the roughing machine 23 several times' to be on the exit side of the roughing mill The rolled web 100 is rolled to the target intermediate strip thickness. Hereinafter, the case of "passing the rolled material through the rolling frame of the roughing machine 23" is referred to as "pass". After rolling by the roughing mill 23, the rolled material enthalpy (10) is transferred from the outlet side of the roughing mill 23 to the inlet side of the finishing mill, and the finishing mill 26 composed of, for example, 5 to 7 rolling stands 260 will be used. The rolled web 100 is rolled to the desired thickness of the product panel. The rolled product taken out from the finishing mill 26 is cooled by a cooling device 27 such as a water cooling device, and then wound up in a coil shape by a coiler 28. The rolls of the rolling stand of the roughing mill 23 are driven by a motor 231, and the rolls of the rolling stand 260 of the finishing mill 26 are driven by a motor 261. Further, a roughing mill inlet side descaler 22' is disposed on the inlet side of the roughing mill 23, and a finishing mill inlet side descaler 25 is disposed on the inlet side of the finishing mill 26. Further, a coil box (coil b〇x;) 24 is disposed in the transfer table area between the roughing mill 23 and the finishing mill 26. The energy consumption prediction device shown in Fig. 1 calculates the predicted value of the energy consumption line 2 322914 201235124 " required for the production of a desired size and quality product. In the following, the detailed explanation of the energy consumption forecast and the detailed operation will be described. The setting calculation means 13 calculates the operation set value AsET using a known model formula based on the operating conditions and the setting calculation Zn«. The workpiece is, for example, the target thickness to be reached on the outlet side of the finishing mill 26, the outlet side temperature of the rolling mill, etc. In addition, the operation setting value is required to be = necessary to make the rolled material 100 a desired thickness The roll gap: and the roll speed required to achieve the desired temperature at the exit of the finishing mill, etc., that is, in order to achieve the desired size and quality of the product produced in the hot rolling line 2 The necessary rolling torque, roll speed, and the like are calculated by the setting calculation means 13. Further, the load limit value of the finishing mill 26 is not exceeded, and the rotation of the motor 261 which does not exceed the roll of the rolling stand 260 is not exceeded. Under the condition of the moment limit value, the setting calculation means 13 calculates the operation setting value ASET including the set values of the rolling load, the rolling torque, and the roll speed of each pass and each rolling stand 260. This special action setting value A 'is best at least: from the bite point which is important for ensuring the accuracy of the front end of the 札 材 丨 丨 、 、 、 、 、 、 、 、 、 Rolling speed of rolled web 100 The large middle point, and to be counted out of the low end of the temperature and rolling material 100 points three points below the calculated value ASET operation setting place is called the "target point." Further, one of the above-described series calculations is referred to as "setting calculation". The operation setting value ASET such as the roll gap and the roll speed calculated by the setting calculation is output to the control device of the hot rolling line 20, and the hot rolling line 20 9 322914 201235124 is operated based on the operation setting value ASET. . In addition, the operation set value AsET of the rolling torque and the roll speed is output to the energy consumption amount calculation device 14 as a parameter for calculating the energy consumption amount. The measured value acquisition means li obtains the rolling length, the roll catching degree, and the rolling length of the entire length of the rolled material 1 in the rolling process from the measuring device (not shown) provided in the hot rolling line 20 Actual measured value aact of rolling power. For example, the rolling torque is calculated using the load applied to the rolls, etc., and the roll speed is calculated using the number of revolutions of the rolls, etc., and the rolling power is calculated using the drive current of the motor 261 or the like. The actual measured value AACT obtained by the actual value acquisition means is output to the setting calculation learning means. The calculation learning device 12 is configured to substitute the actual measured value Aact obtained by the actual value acquisition device 11 into the parameter of the model formula to calculate the actual measured value Aactcal. Further, the setting calculation learning device 12 learns the actual measured value by comparing the actual measured value Aact of each target point of the rolled material 1〇〇 obtained by the actual value acquisition device 11 with the calculated measured value Aactcal.误差α and the error of the measured measured value AAaeAL. Specifically, the setting calculation learning means 12 calculates the ratio of the actual measured value AAeT to the actually measured calculated value AKTGAL. In other words, the setting calculation learning device 12 calculates the set calculation learning value ZnM as the "operating actual measurement value AAeT / the operation actual measurement calculation value AAa (: AL j . The calculated setting calculation learning value 储存 Μ is stored in the set calculation learning value storage device 121 The set calculation learning value Ζη stored in the set calculation learning value storage means 121 is used by the setting calculation means 13. 10 322914 201235124 - Energy consumption and source consumption of the hot rolling line 20 by the energy consumption amount calculating means 14 In the calculation, the operation setting value ASET calculated by the setting calculation device 13 is used. Specifically, the energy consumption amount calculation device 14 sets the operation value ASET and the operation condition based on the calculated rolling torque and the roll speed. To calculate the energy consumption calculation value EnSET of each pass in the roughing mill 23 and each rolling stand 260 of the finishing mill 26. For example, each pass and the motor that drives the rolls of each rolling stand 260 The energy consumption calculation value EnSET of 261 is as shown in the following formulas (1) and (2), by the rolling torque G(t) [kNm] and the roll speed v(t) [m/s Accumulate Time (t) [s] is calculated by integrating:

EnSET= η S Pw(t)dt ---(1)EnSET= η S Pw(t)dt ---(1)

Pv(t)=(1000xv(t)xG(t))/R ...(2) 式(1)中’ $ dt表示從t=0到T,亦即從對於被軋延材loo 的軋延處理開始到結束為止之時間積分,7/係為電力變換 效率(電流-工作間之變換時的效率)^式(2)中,R[mm]為軋 輥半徑’ Pw(t)[kW]為軋延功率。軋延轉矩G(t)係為軋輥, 基準。 能源消費量實際值算出裝置15 ’係使用設定計算學習 裝置12進行設定計算學習值Ζπμ的算出之際所使用之軋延 轉矩的動作實測計算值GiACTCAL及軋輥速度的動作實測計算 值,來算出能源消費量實測計算值EnACTCAL。能源消費 量實測計算值EnACTCAWf、使用式(1)及式(2),藉由以下之式 (3)及式(4)來算出: p^ACTCAL^ (1〇〇〇xViACTCALxGiACTCAL) y R ••⑶ 322914 11 201235124Pv(t)=(1000xv(t)xG(t))/R (2) In the formula (1), '$dt represents from t=0 to T, that is, from rolling for the rolled material loo The time integral from the start to the end of the process, 7/ is the power conversion efficiency (the efficiency at the time of the current-work transition). In the formula (2), R[mm] is the roll radius 'Pw(t)[kW] For rolling power. The rolling torque G(t) is a roll, a standard. The energy consumption actual value calculation device 15' calculates the operation actual measurement calculation value GiACTCAL and the roll speed operation calculation value used for the calculation of the set calculation learning value Ζπμ by the setting calculation learning device 12 Energy consumption measured value EnACTCAL. The calculated energy consumption calculation EnACTCAWf, using equations (1) and (2), is calculated by the following equations (3) and (4): p^ACTCAL^ (1〇〇〇xViACTCALxGiACTCAL) y R •• (3) 322914 11 201235124

EnACKAL= Σ (PwiACKu + Pwi+iAcmL)xSiACT/2 …(4) 式(3)之ί^ΑεΚΑί係為在標的點i之軋延功率的實測計算值, R[mm]為軋輥半徑。式(4)中,SiAGT[sec]為標的點卜i + Ι間 的時間,η為道次或軋延機架260的編號。Σ為表示從最 初的標的點到最後的標的點Μ之總和。 所算出的能源消費量實測計算值EnAeTeu,係輸出至能 源消費量學習值算出裝置17。 能源消費量實際值取得裝置16,係將由實測值取得裝 置11所取得之軋延功率的動作實測值予以積分來算出能 源消費量實際值EnAeT。第3圖顯示能源消費量實際值EnAeT 的算出方法的一個例子。第3圖中,縱軸為軋延功率的動 作實測值PwAeT,橫軸為時間t。 如以下之式(5 )所示,將在測量地點j中之軋延功率的 動作實測值PwAGT(j)與時間步距(time step)Z\t(j)的乘積 予以加算到最終測量地點,就可正確地算出能源消費量實 際值EnAeT :EnACKAL= Σ (PwiACKu + Pwi+iAcmL) xSiACT/2 (4) Equation (3) ί^ΑεΚΑί is the measured value of the rolling power at the point i, and R[mm] is the roll radius. In the formula (4), SiAGT[sec] is the time between the target points i + ,, and η is the number of the pass or the rolling stand 260. Σ is the sum of the points from the initial point of the mark to the last point of the mark. The calculated calculated energy consumption amount EnAeTeu is output to the energy consumption learned value calculation means 17. The energy consumption actual value obtaining means 16 integrates the actual measured value of the rolling power obtained by the actual value obtaining means 11 to calculate the energy consumption actual value EnAeT. Fig. 3 shows an example of a method of calculating the actual energy consumption amount EnAeT. In Fig. 3, the vertical axis is the actual measured value PwAeT of the rolling power, and the horizontal axis is the time t. As shown in the following formula (5), the product of the measured value PwAGT(j) of the rolling power in the measurement point j and the time step Z\t(j) is added to the final measurement location. , the actual value of energy consumption EnAeT can be calculated correctly:

EnACT= S P/CT(j)(t)dt= Z(PwiACT(j)xAt(j)) …(5) 式(5)中,$ dt為表示從t=0到T之時間積分,Σ為表示 從:j=0到L-1之總和。L為最終測量地點。 算出的能源消費量實際值ΕηΑα,係輸出至能源消費量 學習值算出裝置17。 能源消費量學習值算出裝置17,係藉由比較能源消費 量實際值EnAeT及能源消費量實測計算值EnAeTeAl1,來學習能 源消費量實際值ΕηΑΠ與能源消費量實測計算值EnAGTeu之誤 12 322914 201235124 差。具體而言,係如以下之式(6)所示,算出能源消費量實 際值EnAeT之相對於能源消費量實測計算值EnAeTGAL之比,以 作為能源消費量學習值ZnECeuR :EnACT=SP/CT(j)(t)dt= Z(PwiACT(j)xAt(j)) (5) In equation (5), $dt is the time integral from t=0 to T, Σ Indicates the sum from: j=0 to L-1. L is the final measurement location. The calculated actual energy consumption amount ΕηΑα is output to the energy consumption amount learning value calculation means 17. The energy consumption learning value calculation device 17 learns the actual value of the energy consumption ΕηΑΠ and the measured value of the energy consumption amount EnAGTeu by comparing the actual energy consumption value EnAeT and the energy consumption calculation value EnAeTeAl1. 12 322914 201235124 . Specifically, as shown in the following formula (6), the ratio of the actual value of the energy consumption amount EnAeT to the measured value EnAeTGAL of the energy consumption amount is calculated as the energy consumption learning value ZnECeuR:

ZnEcCUR-EnACVEnACTCAL …(6) 算出的能源消費量學習值ZnECeuR係儲存至能源消費量學習 值儲存裝置171。 在能源消費量學習值儲存裝置171中,藉由使用過去 算出之已儲存在能源消費量學習值儲存裝置171之舊的能 源消費量學習值(以下將之表示成「ZnECGLD」)與新算出的能 源消費量學習值ZnECeUR之如式(7)所示之加權平均演算,來 重新算出能源消費量學習值ZnEC。ZnEcCUR-EnACVEnACTCAL (6) The calculated energy consumption learning value ZnECeuR is stored in the energy consumption learning value storage device 171. In the energy consumption learning value storage device 171, the old energy consumption amount learning value (hereinafter referred to as "ZnECGLD") which has been stored in the energy consumption amount learning value storage device 171 calculated in the past is used and newly calculated. The energy consumption learning value ZnECeUR is calculated as the weighted average calculation shown in equation (7) to recalculate the energy consumption learning value ZnEC.

ZnEc=(l—d) ZriEc0LD + ο: ZnEcCUR …(7) 式(7)中,α為加權係數。使用式(7)加以更新之新的能源 消費量學習值ZnEC係儲存至能源消費量學習值儲存裝置 171。 預測值算出裝置18,係使能源消費量學習值儲存裝置 171中儲存的能源消費量學習值Zn.EC反映至由能源消費量 算出裝置14所算出之能源消費量計算值EnSET,而算出能源 消費量預測值EnPfed。具體而言,係使用以下之式(8),來 算出已將能源消費量學習值ZnEC考慮進去之能源消費量預 測值 EnPfed :ZnEc=(l−d) ZriEc0LD + ο: ZnEcCUR (7) In the formula (7), α is a weighting coefficient. The new energy consumption learning value ZnEC updated using the equation (7) is stored in the energy consumption learning value storage device 171. The predicted value calculation means 18 converts the energy consumption amount learning value Zn.EC stored in the energy consumption amount learning value storage means 171 to the energy consumption amount calculation value EnSET calculated by the energy consumption amount calculation means 14 to calculate the energy consumption. The quantity predicted value EnPfed. Specifically, the following equation (8) is used to calculate the energy consumption prediction value EnPfed that has taken into account the energy consumption learning value ZnEC:

EnPred= ZnEcxEnSET …(8) 如上所述,對軋延功率進行時間積分而得到之能源消 費量實際值ΕηΑα,使用從實測值算出之軋延轉矩及計算軋 13 322914 201235124 輥速度的動作實測計算值AAeTC^來算出之能源消費量實測 計算值EnAGTGU,然後比較能源消費量實際值EnAeT及能源消 費量實測計算值EnAeKR來求出能源消費量學習值Zmc,再 使用此能源消費量學習值ZnEC就可高精確度地算出正確的 能源消費量預測值Er^ed。 如以上說明的,本發明第一實施形態之能源消費量預 測裝置10,係使用以軋延轉矩及軋輥速度作為能源消費量 計算的輸入參數之能源消費量計算式,來算出能源消費量 計算值。此時,軋延轉矩及軋輥速度係使用設定計算的模 型公式來算出。 作為能源消費量計算式的輸入參數之軋延轉矩及軋輥 速度的動作實測值AAeT與動作實測計算值AAeTeAL之誤差,係 使用由設定計算學習裝置12所算出之設定計算學習值ZnM 來加以消除。 而且,在能源消費量的學習中,係藉由學習能源消費 量實測計算值EnAeTeAL(使用以軋延轉矩及軋輥速度的動作 實測計算值ΑΑα(^作為輸入參數之能源消費量計算式來算 出者)與能源消費量實際值EnAeT(使用來自熱軋作業線20 之回授資訊之軋延功率的動作實測值來算出者)之誤差,來 消除能源消費量計算式本身的誤差。 因此,根據第1圖所示之能源消費量預測裝置10,將 能源消費量計算式的輸入參數的誤差、與能源消費量計算 式本身的誤差予以區分開來,並就各誤差分別加以學習, 來提高能源消費量的預測精確度。 14 322914 201235124 - (第二實施形態) i 本發明第二實施形態之能源消費量預測裝置10,係如 第4圖所示,在另外具備有能源消費量學習值更新裝置19 之點與第1圖所示的能源消費量預測裝置10不同。其他的 構成都與第1圖所示的第一實施形態相同。 能源消費量學習值更新裝置19,如後面的詳細說明 般,係依據由設定計算學習裝置12所算出之設定計算學習 值乙❿的變化率,將能源消費量學習值儲存裝置171中儲存 的能源消費量學習值ZnEC除以設定計算學習值ZnM的變化 率,來算出新的能源消費量學習值ZnEC。然後,預測值算 出裝置18使用新的能源消費量學習值ZnEC來算出能源消費 量預測值EnPl^d。 能源消費量實際值EnAeT與能源消費量計算值EnSET之誤 差,也會因為在軋延轉矩及軋輥速度的算出中使用的模型 公式的預測誤差而產生。軋延轉矩及軋輥速度的設定計算 學習值ZnM有很大的變化之情況時,作為能源消費量計算的 輸入參數之軋延轉矩及軋輥速度就會變化,所以有能源消 費量的預測精確度降低之虞。 因此,在能源消費量學習計算中,有必要判斷軋延轉 矩及軋輥速度的設定計算學習值Ζ η μ是否飽和。此處所謂的 「設定計算學習值飽和」,係指即使重複進行熱軋作業線中 的軋延處理,設定計算學習值也幾乎不變化之意。例如, 設定計算學習值的變化率在10%以下之情況時,就判斷為 飽和。 15 322914 201235124 若軋延轉矩及軋輥速度的設定計算學習值ZnM並未飽 和,就有必要在接下來要處理之被軋延材100的能源消費 量計算中,考慮到設定計算學習值ZnM的變化份量而將該變 化份量予以除外。 能源消費量學習值更新裝置19,係從設定計算學習值 儲存裝置121將屬於能源消費量計算的輸入參數之軋延轉 矩及軋輥速度的設定計算學習值ZnM予以讀出。讀出的設定 計算學習值ΖηΜ,係為更新前的設定計算學習值及更 新後的設定計算學習值ZnMNEW。其中,由於軋延速度與軋輥 速度具有大致成正比之關係,所以使用軋延速度的設定計 算學習值也沒有任何問題。 為了在能源消費量計算中將設定計算學習值的變化份 量予以除外,而進行以下的處理。亦即,能源消費量學習 值更新裝置19係比較更新前的設定計算學習值ΖηΓβ及更 新後的設定計算學習值Zn/Eff,算出設定計算學習值的變化 率万Μ。當此設定計算學習值的變化率点M為預先設定的一定 的閾值7以上時,就將設定計算學習值儲存裝置121中儲 存的設定計算學習值除以設定計算學習值的變化率/5 m所 得到之值,使用作為適用於能源消費量計算之設定計算學 習值Ζπμ。閾值τ係為例如0. 1。 具體而言,係使用式(9)來算出設定計算學習值的變化 率 β Ά .. y5H=ZnMNEVZnM0LD …(9) 此處,在r引1-冷μ|之情況時,係依照以下之式(10), 16 322914 201235124 -- 來算出適用於能源消費量計算之新的能源消費量學習值 ^ Ziiec :EnPred= ZnEcxEnSET (8) As described above, the actual value of the energy consumption obtained by time-integrating the rolling power ΕηΑα, using the rolling torque calculated from the measured value and calculating the operation of the rolling 13 322914 201235124 roller speed The value AAeTC^ is used to calculate the calculated energy consumption value EnAGTGU, and then the energy consumption actual value EnAeT and the energy consumption measured value EnAeKR are compared to calculate the energy consumption learning value Zmc, and then the energy consumption learning value ZnEC is used. The correct energy consumption prediction value Er^ed can be calculated with high accuracy. As described above, the energy consumption amount predicting device 10 according to the first embodiment of the present invention calculates the energy consumption amount using the energy consumption calculation formula of the input parameter calculated by the rolling torque and the roll speed as the energy consumption amount. value. At this time, the rolling torque and the roll speed were calculated using the model formula of the set calculation. The error between the measured value AAeT of the rolling torque and the rolling speed and the measured value AATeAL of the measured value as the input parameter of the energy consumption calculation formula is eliminated by using the set calculation learning value ZnM calculated by the setting calculation learning device 12 . Moreover, in the study of energy consumption, it is calculated by learning the calculated value of energy consumption EnAeTeAL (using the measured value of the measured value of the rolling torque and the roll speed ΑΑα (^ as the input parameter energy consumption calculation formula) The error of the energy consumption calculation formula itself is eliminated by the error of the energy consumption actual value EnAeT (calculated using the actual measured value of the rolling power from the feedback information of the hot rolling line 20). The energy consumption prediction device 10 shown in Fig. 1 distinguishes the error of the input parameter of the energy consumption calculation formula from the error of the energy consumption calculation formula itself, and learns each error separately to improve the energy. (Embodiment 2) The energy consumption amount prediction device 10 according to the second embodiment of the present invention has an energy consumption amount learning value update as shown in Fig. 4 The point of the device 19 is different from the energy consumption amount predicting device 10 shown in Fig. 1. The other configurations are the same as those of the first embodiment shown in Fig. 1. The energy consumption learning value updating means 19 calculates the energy stored in the energy consumption learning value storage means 171 based on the change rate of the learning value ❿ calculated based on the setting calculated by the setting calculation learning means 12, as will be described in detail later. The consumption learning value ZnEC is divided by the rate of change of the set calculation learning value ZnM to calculate a new energy consumption learning value ZnEC. Then, the predicted value calculating means 18 calculates the energy consumption predicted value using the new energy consumption learning value ZnEC. EnPl^d. The error between the actual energy consumption value EnAeT and the energy consumption calculation value EnSET is also caused by the prediction error of the model formula used in the calculation of rolling torque and roll speed. Rolling torque and roll When the speed is set to calculate a large change in the learning value ZnM, the rolling torque and the roll speed, which are input parameters for energy consumption calculation, change, so the accuracy of prediction of energy consumption is reduced. In the energy consumption learning calculation, it is necessary to judge the setting of the rolling torque and the roll speed to calculate whether the learning value Ζ η μ is full. Here, the "set calculation learning value saturation" means that even if the rolling process in the hot rolling line is repeated, the set calculation learning value hardly changes. For example, the rate of change of the set calculation learning value is set to 10 If it is less than %, it is judged to be saturated. 15 322914 201235124 If the rolling torque and the roll speed are set, the learning value ZnM is not saturated, and it is necessary to process the energy consumption of the rolled material 100 to be processed next. In the calculation, the change amount is set in consideration of setting the change amount of the calculated learning value ZnM. The energy consumption learned value update means 19 measures the input parameter belonging to the calculation of the energy consumption amount from the set calculation learned value storage means 121. The setting of the extension torque and the roll speed calculation learning value ZnM is read. The readout calculation The calculation learning value ΖηΜ is calculated by calculating the learning value before the update and the updated calculation calculation value ZnMNEW. Among them, since the rolling speed is approximately proportional to the roll speed, there is no problem in calculating the learning value using the setting of the rolling speed. In order to exclude the change amount of the set calculation learning value in the energy consumption calculation, the following processing is performed. In other words, the energy consumption learning value update means 19 compares the set calculation learning value Ζ Γ Γ before the update and the set calculation learning value Zn / Eff after the update, and calculates the rate of change of the set calculation learning value. When the change rate point M of the set calculation learning value is a predetermined threshold value 7 or more, the set calculation learning value stored in the set calculation learning value storage means 121 is divided by the change rate of the set calculation learning value / 5 m. The value obtained is calculated using the set value Ζπμ as a setting suitable for energy consumption calculation. The threshold τ is, for example, 0.1. Specifically, the rate of change β Ά of the set calculation learning value is calculated using the equation (9). y5H=ZnMNEVZnM0LD (9) Here, in the case of r-lead 1-cold μ|, the following equation is used. (10), 16 322914 201235124 -- To calculate the new energy consumption learning value for energy consumption calculations ^ Ziiec :

ZllEC = ZriEC〇LD/ Μ …(10) 另一方面,在11 - /3 Μ I〈 7之情況時,係將更新前的能 源消費量學習值ΖπεΛ15直接使用作為適用於能源消費量計 算之能源消費量學習值ZnEc。亦即,Ziiec =ZnEcQIjD。 如上述決定出之能源消費量學習值Zruc,係輸出至預 測值算出裝置18。預測值算出裝置18係算出反映了能源 消費量學習值Ziiec之能源消費量。除以上所述者外,皆與 第一實施形態實質相同,在此省略重複的記載。 如以上所說明的,根據第二實施形態之能源消費量預 測裝置10,就可避免演算能源消費量計算的輸入參數之設 定計算學習裝置12、及能源消費量學習值算出裝置17雙 方重複學習能源消費量計算的輸入參數的預測誤差之情 形。結果,就可使能源消費量預測的精確度穩定及提高能 源消費量預測的精確度。 (第三實施形態) 本發明第三實施形態之能源消費量預測裝置10,係如 第5圖所示,在具備有用來儲存依被軋延材100的板厚、 板寬及鋼種分別加以區分之複數個能源消費量學習值之學 習值資料庫30之點與第一實施形態不同。其他的構成都與 第1圖所示的第一實施形態一樣。 軋延速度、軋輥速度及軋延時間,係依被軋延材100 的板厚、板寬及鋼種而異。因此,能源消費量的預測誤差 17 322914 201235124 會隨者被軋延材刚的板厚、板寬及鋼種之不同而不同。 因而,準備依被軋延材⑽的板厚、板纽_分別加以 區分之各種學習值來作為能源消f量學習值2此後有效。 第6圖顯示學習值資料庫30中儲存的表(table)的一 個例子。此表的構成,係針對各鋼種都準備一個表單 (sheet) ’然後在各表單中以板厚及板寬進行區分,逐一記 錄各區分的能源消費量學習值z nec。 β 例如,以板厚區分丨為^“…至i.ymm]、板寬區分 2為980[mm]至n〇〇[mm]之方式,以決定好的範圍來劃分 板厚及板寬,並對每個區分編列流水編號。然後,在第6 圖所不的表中記錄各個區分之能源消費量學習值。 在熱軋作業線2〇之被軋延材1〇〇的軋延處理後,就將 由能源消費量學習值算出裝置Π所算出的能源消費量學 習值輸入至能源消費量學習值儲存裝置171。此時,與接 觉札延處理的被軋延材100的板厚、板寬及鋼種的區分對 應之能源消費量學習值,係使用作為式(7)中之已儲存的能 源4費量學習值ZnEcQLD。然後,將使用式(7 )加以更新後之 能源消費量學習值以扣當作是新的能源消費量學習值,予 以記錄到學習值資料庫30中儲存的表内之對應的區分。另 方面’將與預定要軋延的被軋延材100的板厚、板寬及 鋼種的區分對應之能源消費量學習值,在軋延處理前輸出 至預測值算出裝置18。 如上所述,在第5圖所示之能源消費量預測裝置1〇 中’係依照被軋延材100的板厚、板寬及鋼種的區分,更 322914 18 201235124 -新各區分的能源消費量學習值並將該值儲存到學習值資料 庫30中。然後,將針對各區分而儲存於學習值資料庫30 中之能源消費量學習值輸出至預測值算出裝置18。預測值 算出裝置18係算出反映了依區分而取得的能源消費量學 習值之能源消費量預測值EnPfed。除以上所述者外,皆與第 一實施形態實質相同,在此省略重複的記載。 如以上所說明的,根據第三實施形態之能源消費量預 測裝置10,具備有每一種板厚、板寬及鋼種的區分之能源 消費量學習值,就可補償依被軋延材100的板厚、板寬及 鋼種的區分而異之預測誤差。結果,就可更正確預測能源 消費量。 (第四實施形態) 本發明第四實施形態之能源消費量預測裝置10,係如 第7圖所示,在另外具備有將預測值算出裝置18所算出的 能源消費量預測值Enpw顯示出來之顯示裝置40之點與第 1圖所示的能源消費量預測裝置10不同。其他的構成都與 第1圖所示的第一實施形態相同。 根據第四實施形態之能源消費量預測裝置10,將所算 出的能源消費量預測值〖^^顯示於顯示裝置40。因此, 操作員及工程師等之熱軋作業線20的作業者,就可經常確 認接下來將處理的被軋延材100的能源消費量。因此,在 能源消費量預測值EnPl_ed很大之情況時,作業者就可視需要 而進行軋延條件的變更。 如上所述,本發明雖藉由第一至第四實施形態而記載 19 322914 201235124 如上,然而不應將作為此揭示的一部份之論述及圖式理解 成用來限定本發明者。透過此揭示,各種代替實施形態、 實施例及運用技術對於本技術領域之業者而言都將變得顯 而易知。亦即,本發明理所當然地包含此處未記載的各種 實施形態。因此,本發明之技術上的範圍只由就上述的說 明來說妥當的申請專利範圍中的發明特定事項所決定。 【圖式簡單說明】 第1圖係顯示本發明第一實施形態之能源消費量預測 裝置的構成之示意圖。 第2圖係顯示熱軋作業線的構成例之示意圖。 第3圖係顯示本發明第一實施形態之能源消費量預測 裝置進行的能源消費量實際值之算出方法的一個例子之示 意圖。 第4圖係顯示本發明第二實施形態之能源消費量預測 裝置的構成之示意圖。 第5圖係顯示本發明第三實施形態之能源消費量預測 裝置的構成之示意圖。 第6圖係顯示本發明第三實施形態之能源消費量預測 裝置的學習值資料庫中儲存之表的一個例子之圖。 第7圖係顯示本發明第四實施形態之能源消費量預測 裝置的構成之示意圖。 【主要元件符號說明】 10 能源消費量預測裝置 11 實測值取得裝置 20 322914 201235124 12 設定計算學習裝置 13 設定計算裝置 14 能源消費量算出裝置 15 能源消費量實際值算出裝置 16 能源濟費量實際值取得裝置 17 能源消費量學習值算出裝置 18 預測值算出裝置 19 能源消費量學習值更新裝置 20 熱軋作業線 21 加熱爐 22 粗軋機入口側去銹皮器 23 粗軋機 24 盤捲箱 25 精軋機入口側去銹皮器 26 精札機 27 冷卻裝置 28 捲取機 30 學習值資料庫 40 顯示裝置 100 被軋延材 121 設定計算學習值儲存裝置 171 能源消費量學習值儲存裝置 260 軋延機架 231 、 261 馬達 21 322914ZllEC = ZriEC〇LD/ Μ ...(10) On the other hand, in the case of 11 - /3 Μ I<7, the energy consumption learning value ΖπεΛ15 before the update is directly used as the energy for energy consumption calculation. Consumption learning value ZnEc. That is, Ziiec = ZnEcQIjD. The energy consumption learning value Zruc determined as described above is output to the predicted value calculation means 18. The predicted value calculation means 18 calculates the energy consumption amount reflecting the energy consumption learned value Ziiec. Except for the above, it is substantially the same as the first embodiment, and the overlapping description is omitted here. As described above, according to the energy consumption amount prediction device 10 of the second embodiment, the setting calculation learning device 12 and the energy consumption amount learning value calculation device 17 which can avoid the input parameter calculation of the calculation of the energy consumption amount are repeatedly learning the energy. The case of the prediction error of the input parameters of the consumption calculation. As a result, the accuracy of energy consumption forecasts can be stabilized and the accuracy of energy consumption forecasts can be improved. (Third Embodiment) The energy consumption amount predicting device 10 according to the third embodiment of the present invention is provided with a thickness, a plate width, and a steel type for storing the rolled material 100 according to the fifth embodiment. The point of the learning value database 30 of the plurality of energy consumption learning values is different from that of the first embodiment. The other configurations are the same as those of the first embodiment shown in Fig. 1. The rolling speed, the roll speed, and the rolling time vary depending on the thickness of the rolled material 100, the width of the sheet, and the type of steel. Therefore, the prediction error of energy consumption 17 322914 201235124 will vary depending on the thickness of the rolled material, the width of the plate and the type of steel. Therefore, it is effective to prepare the learning value 2 as the energy consumption amount 2 depending on the thickness of each of the rolled material (10) and the respective learning values. Fig. 6 shows an example of a table stored in the learning value database 30. The composition of this table is to prepare a sheet for each steel type and then distinguish between the thickness and the width of the sheet in each form, and record the energy consumption learning value z nec of each division. β For example, the thickness of the plate is defined as ^"...to i.ymm], and the width of the plate is 2 to 980 [mm] to n〇〇[mm], and the plate thickness and the plate width are determined by a predetermined range. The serial number is also assigned to each division. Then, the energy consumption learning values of each division are recorded in the table shown in Figure 6. After the rolling treatment of the rolled material of the hot rolling line 2〇 The energy consumption learning value calculated by the energy consumption learning value calculation means 输入 is input to the energy consumption amount learning value storage means 171. At this time, the thickness and the board of the rolled material 100 processed by the sensation is processed. The energy consumption learning value corresponding to the difference between the width and the steel type is used as the stored energy 4 amount learning value ZnEcQLD in the formula (7). Then, the energy consumption learning value updated by using the formula (7) is used. The deduction is regarded as a new energy consumption learning value, and is recorded in the corresponding division in the table stored in the learning value database 30. On the other hand, 'the thickness of the rolled material 100 to be rolled, The difference between the board width and the steel type corresponds to the energy consumption learning value, in the rolling The pre-processing is output to the predicted value calculation device 18. As described above, in the energy consumption amount prediction device 1 shown in Fig. 5, the difference is based on the thickness, the plate width, and the steel type of the rolled material 100. 18 201235124 - The energy consumption learning value of each new division is stored and stored in the learning value database 30. Then, the energy consumption learning value stored in the learning value database 30 for each division is output to the predicted value. The calculation device 18. The predicted value calculation device 18 calculates the energy consumption predicted value EnPfed that reflects the energy consumption learned value obtained by the division. The above is substantially the same as the first embodiment, and is omitted here. As described above, the energy consumption amount predicting device 10 according to the third embodiment is provided with the learning value of the energy consumption amount of each of the plate thickness, the plate width, and the steel type, and the compensation can be compensated. The prediction error of the material thickness, the plate width, and the steel type of the material 100 is different. As a result, the energy consumption can be more accurately predicted. (Fourth embodiment) The fourth embodiment of the present invention As shown in FIG. 7, the consumption prediction device 10 further includes the display device 40 that displays the energy consumption predicted value Enpw calculated by the predicted value calculation device 18, and the energy consumption shown in FIG. The amount prediction device 10 is different. The other configuration is the same as that of the first embodiment shown in Fig. 1. According to the energy consumption amount prediction device 10 of the fourth embodiment, the calculated energy consumption amount prediction value is displayed on The display device 40. Therefore, the operator of the hot rolling line 20 such as an operator or an engineer can often confirm the energy consumption of the rolled product 100 to be processed next. Therefore, the energy consumption predicted value EnPl_ed is very In the case of a large situation, the operator can change the rolling conditions as needed. As described above, the present invention is described above by way of the first to fourth embodiments. The above description and the drawings are not to be construed as limiting the invention. Various alternative embodiments, examples, and operational techniques will become apparent to those skilled in the art from this disclosure. That is, the present invention naturally includes various embodiments not described herein. Therefore, the technical scope of the present invention is determined only by the specific matters of the invention in the scope of the patent application as described above. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a view showing the configuration of an energy consumption amount predicting device according to a first embodiment of the present invention. Fig. 2 is a schematic view showing a configuration example of a hot rolling line. Fig. 3 is a view showing an example of a method of calculating the actual value of the energy consumption amount by the energy consumption amount predicting device according to the first embodiment of the present invention. Fig. 4 is a view showing the configuration of an energy consumption amount predicting device according to a second embodiment of the present invention. Fig. 5 is a view showing the configuration of an energy consumption amount predicting device according to a third embodiment of the present invention. Fig. 6 is a view showing an example of a table stored in a learning value database of the energy consumption amount predicting device of the third embodiment of the present invention. Fig. 7 is a view showing the configuration of an energy consumption amount predicting device according to a fourth embodiment of the present invention. [Description of main component symbols] 10 Energy consumption prediction device 11 Actual value acquisition device 20 322914 201235124 12 Setting calculation learning device 13 Setting calculation device 14 Energy consumption calculation device 15 Energy consumption actual value calculation device 16 Energy consumption actual value Acquisition device 17 Energy consumption learning value calculation device 18 Predicted value calculation device 19 Energy consumption learning value update device 20 Hot rolling operation line 21 Heating furnace 22 Roughing mill inlet side descaling machine 23 Roughing mill 24 Coil rolling box 25 Finishing mill Inlet side descaler 26 Finishing machine 27 Cooling device 28 Coiler 30 Learning value database 40 Display device 100 Rolled material 121 Setting calculation learning value storage device 171 Energy consumption learning value storage device 260 Rolling frame 231, 261 motor 21 322914

Claims (1)

201235124 七、申請專利範圍: 1. 一種能源消費量預測裝置,係熱軋作業線之能源消費量 預測裝置,具備有: 實測值取得裝置,取得在前述熱軋作業線的軋延處 理中所測量的動作實測值; 設定計算學習裝置,比較將前述動作實測值應用至 模型公式的參數而得到的動作實測計算值與前述動作 實測值,而算出設定計算學習值; 設定計算裝置,使用前述熱軋作業線的作業條件及 前述設定計算學習值,來計算包含前述熱軋作業線中的 前述軋延轉矩、前述軋輥速度及前述軋延功率的設定值 在内的動作設定值; 能源消費量算出裝置,使用前述動作設定值來算出 能源消費量計算值; 能源消費量實際值算出裝置,使用前述軋延轉矩及 前述軋輥速度的前述動作實測計算值來算出能源消費 量實測計算值; 能源消費量實際值取得裝置,藉由將前述軋延功率 的前述動作實測值予以積分來取得能源消費量實際值; 能源消費量學習值算出裝置,藉由比較前述能源消 費量實測計算值及前述能源消費量實際值,以算出能源 消費量學習值;以及 預測值算出裝置,算出使前述能源消費量學習值反 映至前述能源消費量計算值而得到的能源消費量預測 1 322914 201235124 值。 2. 如申請專利範圍第1項所述之能源消費量預測裝置,其 中, 前述預測值算出裝置,係使用將過去算出之舊的能 源W肖費里學習值加權至前述能源消費量學習值而得到 的能源消費量學習值,來算出前述能源消費量預測值。 3. 如申請專利範圍第1或第2項所述之能源消費量預測裝 置,其中,復具備有: 能源消費量學習值更新裝置,在前述設定計算學習 =的變化率在一定值以上之情況時,將前述能源消費量 予邊值除以前述設定計算學習值的變化率而算出新的 能源消費量學習值。 如申明專利範圍帛!或第2項所述之能源消費量預測裝 置,其中,復具備有: 干‘值負料庫’用來儲存依前述被軋延材的板厚、 板寬及鋼種分別加以區分之複數個能源消費量學習值。 •如申睛專利乾圍帛1或第2項所述之能源消費量預測裝 置,其中,復具備有: 顯不裝置’用來顯示前述能源消費量預測值。 322914 2201235124 VII. Scope of application for patents: 1. An energy consumption forecasting device, which is a device for predicting the energy consumption of a hot rolling line, which is provided with: a measured value obtaining device, which is measured in the rolling process of the hot rolling line The measured operation value is set; the calculation learning device is set, and the actual measured value obtained by applying the measured value of the motion to the parameter of the model formula and the actual measured value are calculated, and the set calculation learning value is calculated; and the calculation device is set to use the hot rolling Calculating the operation value including the rolling torque, the rolling speed, and the set value of the rolling power in the hot rolling line, the working condition of the working line and the setting calculation learning value; The device calculates the energy consumption amount calculation value using the operation setting value; the energy consumption amount actual value calculation device calculates the measured value of the energy consumption amount using the measured value of the rolling operation and the measured value of the rolling speed; Quantity actual value acquisition device by using the aforementioned rolling power The actual measured value of the operation is integrated to obtain the actual value of the energy consumption; the energy consumption learning value calculation device calculates the energy consumption learning value by comparing the measured value of the energy consumption and the actual value of the energy consumption; and the prediction The value calculation device calculates an energy consumption amount prediction 1 322914 201235124 value obtained by reflecting the energy consumption amount learning value to the energy consumption amount calculation value. 2. The energy consumption amount prediction device according to the first aspect of the invention, wherein the predicted value calculation device weights the old energy W-Shawli learning value calculated in the past to the energy consumption amount learning value. The obtained energy consumption learning value is used to calculate the aforementioned energy consumption prediction value. 3. The energy consumption forecasting device according to the first or second aspect of the patent application, wherein the energy consumption learning value update device is provided, and the change rate of the learning calculation = is more than a certain value in the foregoing setting calculation At this time, the new energy consumption learning value is calculated by dividing the energy consumption amount by the side value by the change rate of the set calculation learning value. Such as the scope of the patent 帛! Or the energy consumption forecasting device according to item 2, wherein the dry-valued negative stock is used to store a plurality of energy sources that are distinguished by the thickness, the width of the sheet, and the steel grade of the rolled material; Consumption learning value. • For example, the energy consumption forecasting device described in Shenkang Patent Co., Ltd. or item 2, in which the device is equipped with: “display device” is used to display the aforementioned energy consumption forecast. 322914 2
TW100109076A 2011-02-18 2011-03-17 Energy consumption prediction device TWI483790B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2011033002A JP5666338B2 (en) 2011-02-18 2011-02-18 Energy consumption prediction device

Publications (2)

Publication Number Publication Date
TW201235124A true TW201235124A (en) 2012-09-01
TWI483790B TWI483790B (en) 2015-05-11

Family

ID=46655126

Family Applications (1)

Application Number Title Priority Date Filing Date
TW100109076A TWI483790B (en) 2011-02-18 2011-03-17 Energy consumption prediction device

Country Status (4)

Country Link
JP (1) JP5666338B2 (en)
KR (1) KR101253409B1 (en)
CN (1) CN102641904B (en)
TW (1) TWI483790B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104968448B (en) * 2013-02-04 2017-07-04 东芝三菱电机产业系统株式会社 The energy-saving control device of roll line
CN103272853B (en) * 2013-05-22 2015-01-28 沈阳工业大学 Device and method for setting rolling reduction and rolling speed of each rack in cold continuous rolling
JP5999265B2 (en) * 2013-08-02 2016-09-28 東芝三菱電機産業システム株式会社 Energy-saving operation recommendation system
CN106413930B (en) * 2014-01-24 2018-07-17 东芝三菱电机产业系统株式会社 The energy consumption prediction meanss of roll line
CN104998913B (en) * 2015-06-29 2017-03-15 东北大学 The Forecasting Methodology of power of motor during a kind of cold rolling
CN105160127B (en) * 2015-09-25 2018-05-01 甘肃酒钢集团宏兴钢铁股份有限公司 Performance prediction method based on CSP flow hot continuous rolling production line Q235B steel grades
KR102045553B1 (en) 2017-12-21 2019-11-18 주식회사 포스코 Apparatus for predicting the energy usage of a heating furnace
CN109647899B (en) * 2018-12-26 2021-06-29 北京科技大学 Method for forecasting power consumption of multi-specification rolled pieces in hot rolling and finish rolling process of strip steel

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS61113719A (en) * 1984-11-07 1986-05-31 Mitsubishi Electric Corp Steel material extracting temperature setting method of heating furnace in hot rolling line
JPH06262223A (en) * 1993-01-14 1994-09-20 Sumitomo Metal Ind Ltd Method for estimating quantity of working electric power in rolling shop
DE19703617A1 (en) * 1997-01-31 1998-08-13 Siemens Ag Drive unit for industrial installations, particularly those producing basic materials
JP3498786B2 (en) * 1999-01-19 2004-02-16 住友金属工業株式会社 Power consumption prediction method
JP2001321810A (en) * 2000-05-10 2001-11-20 Nippon Steel Corp Method for predicting power consumption in rolling plant
JP2002165362A (en) * 2000-11-24 2002-06-07 Sumitomo Metal Ind Ltd Forecasting method of electric energy consumption and control method
JP4119684B2 (en) * 2002-05-27 2008-07-16 東芝三菱電機産業システム株式会社 Learning control device for rolling mill setting calculation device
KR100973898B1 (en) * 2003-08-05 2010-08-03 주식회사 포스코 Mill pacing control method for reheating furnance
KR100995600B1 (en) * 2003-11-04 2010-11-22 주식회사 포스코 Temperature Compensation Apparatus of Hot Rolled Sheet Getting into Finishing Mill Stand and Method Thereof
JP5003483B2 (en) * 2006-07-26 2012-08-15 東芝三菱電機産業システム株式会社 Material prediction and material control device for rolling line
KR101271973B1 (en) * 2006-12-27 2013-06-07 주식회사 포스코 Method for single phase rolling in mimi-mill processing
CN101201871B (en) * 2007-12-18 2010-06-09 东北大学 Rigid plasticity finite element method for estimation of roll-force in hot rolling process
KR101357346B1 (en) * 2009-03-13 2014-02-03 도시바 미쓰비시덴키 산교시스템 가부시키가이샤 Optimizing apparatus

Also Published As

Publication number Publication date
JP2012170962A (en) 2012-09-10
CN102641904A (en) 2012-08-22
TWI483790B (en) 2015-05-11
CN102641904B (en) 2014-07-30
JP5666338B2 (en) 2015-02-12
KR20120095273A (en) 2012-08-28
KR101253409B1 (en) 2013-04-11

Similar Documents

Publication Publication Date Title
TW201235124A (en) Energy consumption prediction device
CN104942019B (en) A kind of cold rolling of strip steel process Automatic control method of width
JPWO2009113719A1 (en) Learning method of rolling load prediction in hot plate rolling.
CN104324948B (en) A kind of rougher of hot strip mill process rolled piece width control method
RU2009117322A (en) METHOD FOR TRACKING THE PHYSICAL STATE OF A HOT SHEET METAL OR A HOT STRIP WITHIN THE CONTROL OF A THICK-SHEET ROLLING MACHINE FOR PROCESSING A HOT SHEET METAL OR HOT
KR101617375B1 (en) Feedforward thickness control method for performance of cold rolling mill
TW200914158A (en) Gauge control apparatus
JP5759206B2 (en) Learning coefficient controller
Liu et al. Algorithm design and application of laminar cooling feedback control in hot strip mill
CN104923568A (en) Control method for preventing strip breaking during thin steel strip cold rolling process
CN102794315B (en) Self-learning method for improving forecasting precision of overall length coiling temperature of strip steel
KR101782281B1 (en) Energy consumption predicting device for rolling line
JP5077437B2 (en) Rolling speed control method for cold tandem rolling mill
JP6136743B2 (en) Method of changing the running plate thickness in cold tandem rolling mills
JP5999265B2 (en) Energy-saving operation recommendation system
JP6358222B2 (en) Pass schedule determination method when changing the thickness of the running plate in continuous cold rolling
JP4109407B2 (en) Method and apparatus for controlling flatness of metal plate
Zhang et al. A new anti-disturbance strategy of automatic gauge control for small workroll cold reversing mill
JP2007203303A (en) Shape control method in cold rolling
CN104801552B (en) Control method for alleviating final rolling temperature fluctuation
JP2009214164A (en) Method of predicting deformation resistance of metal strip and method of setting up cold tandem mill
JP3771781B2 (en) Thick steel plate rolling equipment and thick steel plate rolling method
JP6210944B2 (en) Control device and control method for continuous rolling mill
JP2021181095A (en) Rolling load prediction method, rolling method, hot-rolled steel sheet manufacturing method, and rolling load prediction model generation method
JP2012045616A (en) Method of controlling plate thickness in rolling mill