1259839 九、發明說明： 發明所屬之技術領域 本發明是關於在一種使用至少一種反應器系統的設施 中由一種聚合物溶融液生產一種聚合物的方法，此反應系 統是在聚合物溶融液中進行一種形成聚合物的化學反應， 该反應是由至少一種反應參數所決定，藉由此方法取得聚 合物熔融液的至少一種狀態變數，並依據此種狀態變數來 控制此種反應器系統的至少一個反應參數。 此外，本發明也是關於一種使用至少一種反應器系統 由一種聚合物熔融液生產一種聚合物的設施，其中此種反 應器系統適合在聚合物熔融液中進行一種化學反應，該反 應是由反應參數所決定，它是利―種感測器取得聚合物 熔融液的至少-種狀態變數，並且利用一種與此種感測器 連接的控制單元來轉換數據，藉此可依據此種狀態變數來 控制至少一種反應參數。 此種反應器系統的一種反應參數可表示一種物理白 量，藉此可以決；t在反應器中反應進行的反應平衡，❸ 此種反應的量可以是溫度、壓力和/,一 ^ 種化學的量（例如| 合物成分的濃度和/或其克-分子量）。 ) 精由此種反應參數^ 以定義並控制在反應器系統中的反庫 汉應進仃。聚合物熔融$ 的狀態依據此反應的反應參數而改變， 此聚合物熔融液έ 狀悲疋藉由此聚合物熔融液的狀態變 夂双所決定，例如次j 產物的濃度、聚合度、溫度及黏度等 田 _ 體積圍出-個空間，在此空間中，聚取：：此設施丨 θ物溶融液與次要, 1259839 由聚合物溶融 物和裂解的產物，以及在聚合物生產過程中 液所形成的氣體和蒸氣一起流動。 先前技術 在開頭所提的方法及系統是先前的技術所熟知的 US-A-5 155 184專利中描述利用紅外線光譜儀的吸收 測量來決定聚合物的分子結構，流經反應器的物質和反應 器的排放週期是依據吸收測量的結果來控制。依據專= US-A-5 155 184,此方法適合用於控制一種或更多烯烴或 是乙烯基單體的聚合。 US-A-5 155 184專利中所有依靠光學測量方法的缺點 是透明的區域（例如玻璃窗）必需被建構在測量點的設施的 外壁上，如果沒有此種透明的區域，聚合物熔融液的光學 檢查將無法在設施的體積内進行。此種裝置的缺點是管件 系統的機械強度顯著降低，對某些類型的聚合物而言（特別 是自行放熱的反應性聚合物），此種機械強度的降低是不被 容許的，因為對於自行放熱的反應的情形會有破裂的風險。 另一種測量原理可由US-A-4 448 943專利得知，利用 此種測量原理可藉由一種交變電磁場來決定一種聚合物的 介電常數’因為聚合物生產程序的程序變數被控制，所以 測得的介電常數會接近具有所需組合物的聚合物的介電常 數。除了介電常數外，也可以計算出一種消散因子，並且 用於決定聚合物的狀態。依據US-A-4 448 943專利的解 說’在低於20 kHz的頻率下，介電常數會大大地受到離子 1259839BACKGROUND OF THE INVENTION 1. Field of the Invention This invention relates to a process for producing a polymer from a polymer melt in a facility using at least one reactor system, the reaction system being carried out in a polymer melt A chemical reaction for forming a polymer, the reaction being determined by at least one reaction parameter by which at least one state variable of the polymer melt is obtained, and at least one of such reactor systems is controlled in accordance with such state variables Reaction parameters. Furthermore, the invention relates to a facility for producing a polymer from a polymer melt using at least one reactor system, wherein the reactor system is adapted to carry out a chemical reaction in the polymer melt, the reaction being the reaction parameter It is determined that the sensor senses at least one state variable of the polymer melt and uses a control unit connected to the sensor to convert the data, thereby controlling according to the state variable At least one reaction parameter. A reaction parameter of such a reactor system may represent a physical white amount, thereby determining the equilibrium of the reaction carried out by the reaction in the reactor, and the amount of such reaction may be temperature, pressure and/or chemistry. The amount (for example, the concentration of the constituents and/or its gram-molecular weight). The precision of this reaction parameter ^ to define and control the anti-coin in the reactor system should be advanced. The state of the polymer melt $ varies depending on the reaction parameters of the reaction, and the polymer melt is determined by the state of the polymer melt, such as the concentration of the secondary product, the degree of polymerization, and the temperature. And viscosity, etc. _ volume encirclement - space, in this space, gathering:: this facility 丨θ substance melt and secondary, 1259839 from polymer melt and cracked products, and in the polymer production process The gas formed by the liquid flows together with the vapor. The method and system of the prior art, which is described in the prior art, is described in the prior art. The absorption measurement using an infrared spectrometer determines the molecular structure of the polymer, the material flowing through the reactor and the reactor. The emission cycle is controlled based on the results of the absorption measurements. This method is suitable for controlling the polymerization of one or more olefins or vinyl monomers, according to US-A-5 155 184. A disadvantage of all optical measuring methods in the US-A-5 155 184 patent is that a transparent region (for example a glazing) must be constructed on the outer wall of the facility at the measuring point, if there is no such transparent region, the polymer melt Optical inspection will not be possible within the volume of the facility. A disadvantage of such a device is that the mechanical strength of the pipe system is significantly reduced. For certain types of polymers (especially self-heating reactive polymers), this reduction in mechanical strength is not tolerated because There is a risk of rupture in the case of an exothermic reaction. Another measurement principle is known from the patent US-A-4 448 943, by which an alternating electromagnetic field can be used to determine the dielectric constant of a polymer 'because the program variables of the polymer production process are controlled, so The measured dielectric constant will approach the dielectric constant of the polymer having the desired composition. In addition to the dielectric constant, a dissipation factor can also be calculated and used to determine the state of the polymer. According to the US-A-4 448 943 patent, at a frequency below 20 kHz, the dielectric constant is greatly affected by ions 1259839
Si》n ’可以制此種頻率範圍來控制聚合物炫 二定的組合物。依據此發表内容，頻率介於20 kiiz到1 Z之間時’此物質的介電常數達到—個^的數值。 、±A 5 2G8 544專利中描述—種環形的介電感測器和 連3式的測里方法’其中聚合物的黏度可以依據介電 ^月因子來決定。此種感測器產生具有介於〇 5Hz到2〇〇 5之間的頻率的電磁交變場，而且只適用於最大内徑為8 么刀的導管，㈣此種管徑對大規模的公司而言是太小了。 大體來"兄’在聚合物生產的領域中，特別是在生產聚 口物的領域中，沒有耐用、準確且可靠的方法及控制系統。 然而，尤其使用現代化生產聚合物的方法有必要將反 ，條件保持特別穩定並且固定在特定的值。這不只是生產 #均—的產物所必要的’也能抑制因為設定的反應條件 超出，過低而發生不想要的次要反應，此種次要反應不只 對產品品質及產品產率有影響，也需要特別的測量來保持 =方法的環保及不具污染性。尤其因為生產方法的環保觀 念愈來愈重要’所以精確地維持反應條件是特別重要的。 發明内容 根據本發明，本案開頭所提的方法達成了上述目標， 2中聚合物熔融液所產生的柯氏力被用於當作至少一=狀 態變數，並用於控制至少一種反應參數。 本發明所謂的’，聚合物熔融液”是指依據本發明的系統 所產生的所有液體，它們可以包含寡聚合物和/或聚合物， 1259839 '才邛刀的4些混合物也可以被稱為溶液。 對本案開頭所提的設施， 法是楹徂一 #, σ本發明達成此目標的方 楗ί、種柯氏力的感測器，並且接# ^ 66 4-Τ ^ ^ （、一種適合依據測 于勺柯氏力進而控制至少一 ^ 7里欠應參數的控制器。 兩種達成目標的答案都很簡單，： ^ ^ 门平囚為柯氏力感測器已 經被用於聚合物生產的設施。 〜本發明，由柯氏力感測器所取得的柯氏力被用於 0有關聚合物_液狀態的資訊’ i用它來控制反應器 糸統中的反應’可以藉由評估柯氏力瞬間的值和/或評估柯 :力隨時間的變化來達成。隨時間變化的評估是注意在一 段寺間内柯氏力如何發展，例如是否發生突然間的改變及 暴增，或是是否連續緩慢地降低。柯氏力隨時間的變化在 下面被稱為柯氏力的“時間軌跡“，可做為計算第一及其他 短暫區間所得的例子。 本發明利用柯氏力感測器來控制在聚合物生產期間的 反應，基本上它與在生產聚合物或三元纖維素混合物的加 工中使用柯氏力感測姦的已知方法（例如在F反2 $ 2 1 l· 7 5 A1 和 DE 100 44 491 A1 或 DE-A-199 49 726 和 WO-A-01 25517中所描述的）是不同的，在這些前案中，柯氏力感測 器並非真正用於控制反應，反而只是用於控制聚合物或纖 維素混合物的組合物，其一部份進入溶液但未進行反應。 FR 2 821 175 A1是關於一種用於控制產物性質的方 法，例如是合成或聚加成的產物，此方法是藉由一種數學 模型的協助來控制，此種數學的模型可複製產物性質與獨 1259839 特的程序變數之間的關係，例如 W如由流變儀、光譜儀或超音 波的方法來決定熔融液的流動性。在FR 2 82i ”5Ai中也 描述由-種柯氏感測器取得—種反聚合試劑的產量。 因此相較於本發明，這些刊物中並未提及利用聚合物 熔融液所產生的柯氏力來控制至少一種反應的參數。fr 2 821 175 A1真正只有描述由柯氏感測器來取得反應過程中 反聚合試劑的產量 在DE 100 44 491 A1中描述使用一種柯氏測量槽來決 定-種充氣且在壓力下的液體（特別是一種可流動的塑性 體）密度的一種密度測量裝置。另外，DE1〇〇 44 49i A論及 —種用於取得並控制液體的氣體充填的裝置，其中充氣測 量裝置包含密度測量裝置。 DE 100 44 491 A1取得並控制在壓力下的充氣流液體 (例如塑膠泡沫）的組合物，然而並沒有化學反應發生也沒 有控制此反應，因此在DE 100 44 491 A1中並未揭示由至 > 一種反應參數來決定一種利用柯氏感測器的輔助來監控 及控制生成聚合物的化學反應。 在WO-A-01 25517中，三元纖維素溶液的密度及質量 流量是由一種柯氏力感測器所得到，纖維素溶液的水含量 疋依據密度而決定，並且啟動氧化胺進料管線的閥門。 DE-A-199 49 725的具體實例也描述纖維素、水及胺氧 化物的二元纖維素溶液的密度是使用一種柯氏力感測器來 /則1，密度是控制纖維素溶液組合物的兩種狀態變數中的 ~種〇 1259839 DE-A-199 49 726 及 WO-A-Ol 255 17 中的柯氏力感測 器不被用於生產一種聚合物也不被用於控制一種化學反應 器的反應，而是用於控制一種不反應的溶液，其中所含的 纖維素已經是一種聚合物，不再需要被製造。 依據柯氏力的測量或是代表柯氏力的一種狀態變數， 在一個至少一種反應器系統中，當至少一種反應參數改變 時，反應平衡合偏移或改變。隨著反應平衡的偏移，在反 應改變期間反應器系統中所產生的反應產物的量及濃度會 改變。此外，可能產生新的反應產物或可能會防礙反應產 物的形成。此種改變也可以是聚合度的改變，例如在一種 生成長鏈聚酯的聚縮合反應中，聚合度也可以藉由柯氏力 感測器來決定’當鏈長增加時，聚合物熔融液的黏度會增 加並影響柯氏力。 例如當作反應參數時，可以控制各種物質之間或單體Si"n' can be used to control such a range of frequencies to control the composition of the polymer. According to this publication, when the frequency is between 20 kiiz and 1 Z, the dielectric constant of this substance reaches a value of ^. The method of ±A 5 2G8 544 describes a ring-shaped dielectric detector and a method of measuring the type 3, wherein the viscosity of the polymer can be determined according to the dielectric constant factor. This type of sensor produces an electromagnetic alternating field with a frequency between 〇5Hz and 2〇〇5, and is only suitable for catheters with a maximum internal diameter of 8 knives, (iv) such a diameter for large-scale companies It is too small. In general, there is no durable, accurate and reliable method and control system in the field of polymer production, especially in the field of producing agglomerates. However, in particular, it is necessary to use a modern method for producing a polymer, which is necessary to keep the conditions stable and fixed at a specific value. This is not only necessary to produce the product of #均—, but also suppresses the occurrence of unwanted secondary reactions because the set reaction conditions are exceeded and too low. This secondary reaction not only affects product quality and product yield, Special measurements are also needed to keep the method environmentally friendly and non-contaminating. Especially because the environmental protection concept of production methods is becoming more and more important, it is particularly important to maintain the reaction conditions accurately. SUMMARY OF THE INVENTION In accordance with the present invention, the method set forth at the outset achieves the above object, and the Coriolis force generated by the polymer melt in 2 is used as at least one = state variable and is used to control at least one reaction parameter. The term "polymer melt" as used in the present invention refers to all liquids produced by the system according to the present invention, which may comprise oligopolymers and/or polymers, and some mixtures of 1259839 'knives may also be referred to as Solution. For the facilities mentioned at the beginning of this case, the method is 楹徂一#, σ The invention achieves the goal of the square 楗, the kind of Coriolis sensor, and # ^ 66 4-Τ ^ ^ (, a It is suitable for controllers based on the Coriolis force measured in the spoon and then controlling at least one of the 7 ill parameters. The two answers to the goal are simple: ^ ^ The door is prisoner for the Coriolis force sensor has been used for aggregation The facility for the production of materials. ~ The present invention, the Korotk force obtained by the Kodak force sensor is used for 0 information about the state of the polymer _ liquid 'i used to control the reaction in the reactor system' can be borrowed It is achieved by assessing the instantaneous value of Coriolis force and/or evaluating the change in K: force over time. The assessment over time is to pay attention to how the Coriolis force develops in a temple, such as whether sudden changes and sudden increases occur. , or whether it is continuously and slowly reduced. The change in time is referred to below as the "time trajectory" of Coriolis force, and can be used as an example for calculating the first and other short intervals. The present invention utilizes a Korotk force sensor to control the reaction during polymer production, Basically it is known for the use of Kodak's force in the processing of polymers or ternary cellulose blends (for example in F counter 2 $ 2 1 l · 7 5 A1 and DE 100 44 491 A1 or DE -A-199 49 726 is different from that described in WO-A-01 25517. In these previous cases, the Coriolis force sensor was not really used to control the reaction, but instead was used to control the polymer or fiber. A composition of a mixture of substances, one part of which enters the solution but does not react. FR 2 821 175 A1 relates to a method for controlling the properties of a product, such as a synthetic or polyaddition product, by a mathematical method Controlled by the model, this mathematical model replicates the relationship between the nature of the product and the program variables of the unique 1259839, such as the rheometer, spectrometer, or ultrasonic method to determine the fluidity of the melt. FR 2 82i ”5Ai The production of a reverse polymerization reagent obtained by a Kodak sensor is also described. Therefore, compared with the present invention, there is no mention in the publications that the Korotk force generated by the polymer melt is used to control at least one reaction. The parameters of fr 2 821 175 A1 are really only described by the Korotkoff sensor to obtain the yield of the reverse polymerization reagent during the reaction described in DE 100 44 491 A1 using a Kodak measuring cell to determine the type of aeration and under pressure A density measuring device for the density of a liquid, in particular a flowable plastomer. In addition, DE 1 〇〇 44 49 i A relates to a device for obtaining and controlling a gas filling of a liquid, wherein the gas filling measuring device comprises a density measuring device Device. DE 100 44 491 A1 obtains and controls a composition of an aerated flow liquid (for example a plastic foam) under pressure, however no chemical reaction takes place and the reaction is not controlled, so that it is not disclosed in DE 100 44 491 A1. A reaction parameter determines a chemical reaction that uses a Kodak sensor to monitor and control the formation of a polymer. In WO-A-01 25517, the density and mass flow rate of the ternary cellulose solution is obtained by a Coriolis force sensor, the water content of the cellulose solution is determined by the density, and the amine oxide feed line is activated. Valve. Specific examples of DE-A-199 49 725 also describe the density of the binary cellulose solution of cellulose, water and amine oxides using a Coriolis force sensor, then the density is controlled by the cellulose solution composition. The Coriolis force sensor in the two state variables of 1259839 DE-A-199 49 726 and WO-A-Ol 255 17 is not used to produce a polymer nor is it used to control a chemical. The reaction of the reactor is used to control a non-reactive solution in which the cellulose contained is already a polymer and no longer needs to be manufactured. Depending on the measurement of Coriolis force or a state variable representative of the Coriolis force, in at least one reactor system, the reaction equilibrium shifts or changes when at least one of the reaction parameters changes. As the equilibrium of the reaction shifts, the amount and concentration of the reaction product produced in the reactor system changes during the reaction change. In addition, new reaction products may be produced or may hinder the formation of the reaction product. Such a change may also be a change in the degree of polymerization. For example, in a polycondensation reaction for producing a long-chain polyester, the degree of polymerization may also be determined by a Korotk force sensor. 'When the chain length is increased, the polymer melt is added. The viscosity will increase and affect the Coriolis force. For example, when used as a reaction parameter, it is possible to control various substances or monomers.
一個聚合物鏈的反應，另外也會影響黏度 聚縮合反應所生成水份由聚合物熔融液的移除 皆由柯氏力感測器來控 個單體或聚合物鏈與另 響黏度，因為它會影響 1259839 本發明的一種特別有益的改善是測得的柯氏力和聚合 物熔融液的狀態或組合物之間的關係不需要預先以分析性 地檢查，而是可以以一種點對點的校正方法利用各種被控 制的聚合物熔融液狀態依據實驗來決定。由各自的測量點 可以定義出一種特性攔位，它特定地將某些聚合物熔融液 的狀態歸屬於測得的柯氏力。使用此種特性欄位可以基於 純經驗的測量來控制反應器系統，不需要知道測得的狀態 變數、柯氏力及反應參數之間的理論關係。此種方法的優 點是藉由柯氏力來取得狀態及控制反應可成功地被運用在 許多完全不同的生產方法。 在控制單元内，可以各種不同的方法實施特性攔位。 第-種可能性是將特性攔位儲存成為參照表，丨中所測得 的一種柯氏力或所測得的一種柯氏力的時間執跡被關聯至 反應系統中所設定的某些反應參數，如果測得的柯氏力落 在參照表的數值之間，則此表可以被内插。 另一種可能性是將控制單元中的特性欄位儲存成一種 =、經適用於測量點的内插的方程式，例如是―種多項式或 :-種傅立葉級數。利用此種方程式’反應器系統中被設 定的反應參數可以被特定地歸屬於所測量的柯氏力。 是使用測量點來訓練出一種電腦執行的 面柯氏力的測量點做為訓練用的數值， 類 另一種可能性 神經網路，一方 另方面，基於所使用的取得狀態，為了得到想要的聚合 物熔融液的狀恶’在反應器1統中那些反應參數必需被設 定。 1259839 /當然可以使用一種簡單的PID控制器，其中在反應器 系、、充中的反應參數是依據測得的柯氏力與參考的柯氏力之 間的誤差來定’例如測量的柯氏力可以與先前所決定的 參考數值(代表聚合物熔融液的參考狀態)來比較，反應器 系統可以根據比較的結果來控制。 當然不只是所設定的反應參數可以藉由不同形式的特 性欄位來具體說明其絕對值’而且目前能被設定的反應參 數的變化更可以具體地以相對值來說明。 此種反應器系統尤其可以依據柯氏力來控制，因此在 反應器系統中聚合物熔融液的至少一種成分的比例是依據 所述的柯氏力而改變，當聚合物的密度及黏度受此種成分 影響時特別有利。與這種活性成分有關的聚合物的組合物 特別容易被發現使用柯氏力測量。 因為聚合物熔融液的黏度及密度會隨溫度變化，如果 設施體積中或測量體積中的聚合物熔融液的溫度被測量並 且被考慮用於控制反應器系統，則反應器系統控制的準確 性可以進一步的提高。例如以一種簡單的方式來實行，藉 由實驗地（或分析地）方法來決定溫度對柯氏力測量的^ 響’並且/或是將溫度導入特性攔位當作另—種用纟控制S 量。另—種T能性是利用一種補償的方程S來補償因為溫 度所引起的柯氏力的變化，在這種補償方程式中考慮到溫 度對柯氏力的影響。 此種控制系統的構成具有優勢，所以可以適合現行的 設施，特別是此種控制系統的測量單元可以被裝配成I一 12 1259839 種元件，元件中的至少一種感測器被單一地結合在現行的 設施上。另外，此種控制系統適合與一種已經存在的柯氏 力感測裔連接’因此現行的設施完全不需修改或/、疋稱微 修改。 本發明的方法及測量系統可以用於例如由聚合物溶融 液生產聚合物或是用於改變一種聚合、聚加成或是聚縮合 過程中成分的濃度、末端基或是聚合度。 在由酯化反應形成聚合物的反應器的後面使用此種方 法及測量系統是有利的，此種酯化反應是在高於200°C的 溫度下由芳香族及脂肪族的二羧酸或二羧酸酯與二醇類或 羥基酸形成聚酯（例如對苯二甲酸或二甲基對苯二甲酸酯 與乙二醇、1，2-，1，3-丙二醇或ι，4-丁二醇形成聚酯），由雙 酚及二苯基以及二芳基碳酸酯形成聚碳酸酯，或是由芳香 族二酸與芳香族二酚形成聚丙烯酸酯，在一種固定的操作 模式下藉由接近達到熱力學的極限可達成產率的提高。 本發明的方法及裝置可以發現的其他優點是可用於追 蹤同時以熔融液和以溶液或以懸浮方式進行聚合反應過程 中的反應進度或是產物的狀態，在此情形中，柯氏力的感 測器可以特別放置於（優選地是立即放置於）各別反應器系 統之後。如果在酯化、酯交換或聚縮合反應器之後使用柯 氏力感測器，則在這些測量點的聚合物尚未完全被轉換， 所以藉由控制酯化、酯交換及聚縮合反應器系統的反應參 數，可以對進入聚縮合的聚合物熔融液進行一種修正的或 後調整的干預。 13 1259839 在這些聚縮合反應器中會進一步進行聚合到所要的最 終產物’尤其是聚合物最終的黏度可使用柯氏力感測器來 監控，一旦有偏差的情形，可以隨意地干預反應器的控制。 例如聚乙烯對苯二甲酸酯的生產中，當最終的黏度下降並 低於參考值時，可以減低聚縮合反應器的氣態空間的壓力 以增加聚合物熔融液中水分的移除，進而促進更長鏈的聚 乙烯對苯二甲酸酯的形成。 另一種柯氏力感測裔可以放置於一種回流管柱上，以 便能平衡回流量並且考慮到聚合物的進料量。 實施方式 以下將參照圖表來舉例說明本發明的單一具體實例， 如果在特定的使 此具體實例的 因為本發明從以上一般性描述清楚易懂，如 用中與這些特色相關的優點不是很重要時， 單一特色可以省略。The reaction of one polymer chain also affects the water formed by the viscosity polycondensation reaction. The removal of the polymer melt is controlled by a Kodak force sensor to control the monomer or polymer chain and the other viscosity, because It can affect 1259839. A particularly beneficial improvement of the present invention is that the relationship between the measured Coriolis force and the state or composition of the polymer melt does not require prior analytical inspection, but can be corrected in a point-to-point manner. The method utilizes various controlled polymer melt states to be determined experimentally. From each measurement point, a characteristic stop can be defined which specifically assigns the state of certain polymer melts to the measured Coriolis force. The use of this property field allows control of the reactor system based on purely empirical measurements without the need to know the theoretical relationship between the measured state variables, the Coriolis force and the reaction parameters. The advantage of this approach is that Coriolis's ability to achieve state and control reactions can be successfully applied to many completely different production methods. Within the control unit, the characteristic block can be implemented in a variety of different ways. The first possibility is to store the characteristic barrier as a reference table, and a Coriolis force measured in the enthalpy or a measured time deviation of a Coriolis force is associated with some of the reactions set in the reaction system. Parameter, if the measured Coriolis force falls between the values in the reference table, the table can be interpolated. Another possibility is to store the characteristic fields in the control unit as an =, an interpolation equation applied to the measurement points, for example a polynomial or a - Fourier series. The reaction parameters set in the reactor system using such an equation can be specifically attributed to the measured Coriolis force. It is the use of measurement points to train a computer-executed surface Coriolis force measurement point as a training value, another possibility of neural network, one side, based on the state of use used, in order to get the desired The reaction parameters of the polymer melt must be set in the reactor system. 1259839 / Of course, a simple PID controller can be used, in which the reaction parameters in the reactor system and charge are determined according to the error between the measured Coriolis force and the reference Coriolis force. The force can be compared to a previously determined reference value (representing the reference state of the polymer melt) and the reactor system can be controlled based on the results of the comparison. Of course, not only the set reaction parameters can be specified by different types of characteristic fields, but also the change of the reaction parameters that can be set at present can be specifically described by relative values. Such a reactor system can be controlled in particular according to the Coriolis force, so that the proportion of at least one component of the polymer melt in the reactor system is varied according to the Coriolis force, when the density and viscosity of the polymer are affected by this It is particularly advantageous when the ingredients are affected. Compositions of polymers associated with such active ingredients are particularly susceptible to being found using Coriolis force measurements. Since the viscosity and density of the polymer melt will vary with temperature, if the temperature of the polymer melt in the facility volume or in the measurement volume is measured and considered for controlling the reactor system, the accuracy of the reactor system control can be Further improvement. For example, in a simple manner, the temperature (or analytically) method is used to determine the temperature of the Korotk force measurement and/or the temperature is introduced into the characteristic block as another type of control S the amount. Another type of T-energy is the use of a compensated equation S to compensate for the change in Coriolis force due to temperature, in which the effect of temperature on the Coriolis force is taken into account. The construction of such a control system has advantages, so it can be adapted to the existing facilities. In particular, the measuring unit of such a control system can be assembled into I-121259839 components, at least one of which is singly integrated in the current On the facility. In addition, such a control system is suitable for connection to an existing Coriolis sensing sensation. Thus the current facility does not require modification or/or nickname micro-modification at all. The method and measurement system of the present invention can be used, for example, to produce a polymer from a polymer melt or to modify the concentration, end group or degree of polymerization of a component during polymerization, polyaddition or polycondensation. It is advantageous to use such a method and a measurement system behind a reactor which forms a polymer from an esterification reaction which is derived from aromatic and aliphatic dicarboxylic acids or at temperatures above 200 ° C. Dicarboxylates form polyesters with glycols or hydroxy acids (eg, terephthalic acid or dimethyl terephthalate with ethylene glycol, 1,2-, 1,3-propanediol or iota, 4- Butanediol forms a polyester), a polycarbonate formed from bisphenol and diphenyl and a diaryl carbonate, or a polyacrylate formed from an aromatic diacid and an aromatic diphenol in a fixed mode of operation An increase in yield can be achieved by approaching the limit of thermodynamics. Other advantages that can be found in the method and apparatus of the present invention are that it can be used to track the progress of the reaction or the state of the product during the polymerization process in the form of a melt and a solution or suspension, in which case the feeling of Coriolis force The detector can be placed, in particular, immediately after being placed in the respective reactor system. If a Kodak force sensor is used after the esterification, transesterification or polycondensation reactor, the polymer at these measurement points has not yet been completely converted, so by controlling the esterification, transesterification and polycondensation reactor systems The reaction parameters allow for a modified or post-adjusted intervention in the polymer melt entering the polycondensation. 13 1259839 In these polycondensation reactors, further polymerization is carried out to the desired final product. In particular, the final viscosity of the polymer can be monitored using a Kodak force sensor. If there is a deviation, the reactor can be optionally intervened. control. For example, in the production of polyethylene terephthalate, when the final viscosity decreases and is lower than the reference value, the pressure in the gaseous space of the polycondensation reactor can be reduced to increase the removal of moisture in the polymer melt, thereby promoting Formation of longer chain polyethylene terephthalate. Another Kodak sensor can be placed on a return column to balance the return flow and allow for polymer feed. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following is a singular example of a specific embodiment of the present invention, which is to be understood by the following general description of the invention. , a single feature can be omitted.
一單體2)與乙 私二甲酸酯PET的設施 的莫耳比進行反應，並 是在一個酯化反應器5 〇〇 mbar 到 0.5 mbar 例如在DE-A-3 544 6傳送到酯交換反 聚縮合的反應器8 14 1259839 及9。聚合物熔融液由最後的預聚縮合反應器9傳送到最 終的反應器或是環形盤狀的反應器1 〇，並得到最終的產物 π (聚乙烯對苯二甲酸酯的聚合物）。藉由真空產生設備1 3 將次要產物由反應器系統8，9，10經由排放管12吸出， 再經由收集管線14傳送到一個裂解產物的精餾丨5，並以 裂解產物1 6離開此設施。佔裂解產物最大部分的酯化反應 器5中的裂解產物也是經由排放管線1 7傳送到裂解產物 的精餾1 5。經由精餾將最初的起始原料分離並且經由一個 回流管1 8再次傳送到酯化反應器5。 由圖1也可以看出測量柯氏力的測量系統i 9 及22依聚合物熔融液流動的方向被分別放置於反應器系 統5 ’ 7 ’ 8及9之後。另一個柯氏力測量系統23被安置在 由裂化產物的精餾1 5到酯化反應器系統5之間的回流管線 上。測里系統19-23可以使用如Krohne的Optimass MFS 7000型或Emerson Process的20〇〇系列的柯氏力感測器。 設施中所用的柯氏力測量系統的數目並不重要，使用 了五個測1系統只是當作例子；依據本發明，在設施1中存 在至少一種19-23形式的系統。 代表作為產物的狀態變數的柯氏力的信號由測量系統 19-23經由數據線路24或數據轉換區域傳送到—個控制系 統25 ’並藉由兩方向作用的控制線路26來控制反應器系 統：’7，’13’15。反應器系統中所設定的或是主要的反 應芩數是藉由合適的感應器（圖 I口 1並未顯不）來取得，並經 由控制線路26傳送到控制系％ “工制糸統25。控制系統25可以使用 15 1259839 種具有處理為及§己憶體的微電腦。 一種用於數據轉換的估算系統27也同樣與控制系統 25連接。使用手動的最終產物的產物分析（以箭頭μ表 示），依據柯氏力測量系統的信號並依據反應器系統中所測 得的反應參數，在估算系統27中計算出一種特徵值Μ， 藉由此種特徵值的辅助讓控制系統25控制反應器系統。 圖1中顯示一個特性攔位的特徵值29的例子，它顯示 的形式是由分離的校正點30所形成的一個區域，其中傳送 到反應系統5的觸媒4的量[以質量百分率（m_% ΚΑτ)表 不]被歸屬於反應器系統5中所測得的溫度A以及由柯氏 力測量系統1 9所測得的柯氏力ρ c。 特性欄位的這種形式的特徵值29可以經由估算系統 27針對每-個柯氏力測量系統（19到22)以及各別歸屬的反 應器系統（5，7 - 10 ’ 13 ’ 15)計算出來。為了做為反應器 系統的此種單一控制的一種替代或是一種補充，可以產生 一種多維的特徵值29,但無法舉例說明，這是一種超曲面， 匕與所有測量系統所測得的柯氏力以及一些或所有反應器 系統所設定的反應參數有關。利用這種形式的控制，不同 反應器系統的反應參數被合併成一種單一的參數場，此種 參數場同時被狀態變數所控制。相較於歸屬於柯氏力的各 別反應器的分離控制，此種形式的控制的優點也就是依據 不同反應器系統的反應參數來考慮。可以以此種方法來控 制的反應參數的例子有在反應系統5中的第一單體、第一 單體及觸請刪，以及每一個反應器系統中的壓力： 16 1259839 停留時間及溫度。 以下以依據圖1的設施來生產聚碳酸酯為基礎，舉例 描述特徵值29的產生： 依據圖1的測試設施所得的最佳值被加在表1中，首 先是不使用柯氏力測量並維持穩定的連續操作，反應參數 及產物的特徵值被儲存在控制系統中。 之後開啟柯氏力測量系統並且將設施的反應條件保持 固定，然後以計算且計劃完善的方式特定地改變反應器系 統中的反應條件，以便追蹤它們對柯氏力測量的影響並且 儲存它們的效應。在此方法中每一個柯氏力測量系統都用 相同的步驟，直到在设施i的操作過程所預期的眾多狀態 參數中計算出特徵值。 々聚奴酸酯（PC)是由當作第一單體的雙酚A (BpA)和當 作第_早體的二苯基碳酸酯（Dpc)以及當作觸媒的酚化鈉 產生。在設施i的操作中，所有的反應器系統在固定的 又—口疋的產里、壓力以及固定的觸媒濃度下讓兩種單 體的莫耳比在〇.9到2.Q之間做逐步的改變，在每—個步驟 t變之^持料所設定的莫耳Μ料相大約兩個小 :柯IΓ在過渡相或是保持相中都要測量柯氏力感測器上 並且歸屬於反應系統中的反應參數。這個方法 中測量 的特被值29的一個初始校正點30，其 的。圖2、冑，氏力與反應器系統的反應參數是互相關聯 莫耳比，±…由柯氏力感測器所測得的柯氏力如何對應到 、 測得的柯氏力可以發現莫耳比。 17 1259839 &、耳比之後’剩餘的反應參數在以的 ()下進行改變，例如在每—種情況中有十個步驟。、心 酉曰化反應器的溫度可以在200吒到35〇〇c之間變化且壓 在2ba“"0mbar之間變化，在穩定狀態下可以得到： 數值的士50〇/〇的產量以及—5〇%到+2〇〇%的觸媒濃度，在改變 反應參數的期間，剩餘的反應參數保持在穩定操作= 值。 由這些變化的設定量可以得到大量的校正點，同時也 得到特徵值29,其中反應參數與柯氏力感測器所取得量是 互相關聯的。 在此種設施的操作中，特徵值29可被用於當作一種多 維查找表，利用一組測得的柯氏力來查詢，反應參數的改 變必需被設定以便達到想要的穩定狀況。 另外’特徵值也可用於訓練類神經網路，以測得的柯 氏力當作輸入向量並以所設定的反應參數達到設定狀態的 改變當作輸出向量。此種類神經網路的功能就如RiUer及a monomer 2) reacts with the molar ratio of the ethyl dicarboxylate PET facility and is transferred to the transesterification in an esterification reactor from 5 〇〇 mbar to 0.5 mbar, for example at DE-A-3 544 6 Reverse polycondensation reactor 8 14 1259839 and 9. The polymer melt is transferred from the final prepolymerization condensation reactor 9 to the final reactor or to the annular disk reactor 1 to obtain the final product π (polyethylene terephthalate polymer). The secondary product is aspirated from the reactor system 8, 9, 10 via the discharge pipe 12 by means of a vacuum generating device 13 and then passed to a rectification crucible 5 of a cracking product via a collection line 14 and exits with the cracked product 16 facility. The cracked product in the esterification reactor 5, which accounts for the largest portion of the cracked product, is also sent to the rectification 15 of the cracked product via the discharge line 17. The initial starting material is separated via rectification and passed again to the esterification reactor 5 via a reflux line 18. It can also be seen from Fig. 1 that the measurement systems i 9 and 22 for measuring the Coriolis force are placed after the reactor systems 5' 7 ' 8 and 9 respectively depending on the direction in which the polymer melt flows. Another Coriolis force measurement system 23 is placed on the return line between the rectification 15 of the cracked product and the esterification reactor system 5. The Measure System 19-23 can use a 20-inch series of Coriolis force sensors such as Krohne's Optimass MFS 7000 or Emerson Process. The number of Korotk force measurement systems used in the facility is not critical, and the use of five test 1 systems is by way of example only; in accordance with the present invention, at least one system of the 19-23 format is present in facility 1. The signal representing the Coriolis force as the state variable of the product is transmitted by the measurement system 19-23 via the data line 24 or the data conversion region to a control system 25' and the reactor system is controlled by a two-way control line 26: '7, '13'15. The set or main reaction parameters set in the reactor system are obtained by a suitable inductor (not shown in Figure 1) and transmitted to the control system via control line 26. "Working System 25 The control system 25 can use 15 1259839 microcomputers with processing and § memory. An estimation system 27 for data conversion is also connected to the control system 25. Product analysis using manual final products (indicated by arrow μ) A eigenvalue Μ is calculated in the estimation system 27 based on the signal of the Korot's force measurement system and based on the measured reaction parameters in the reactor system, with the aid of such eigenvalues allowing the control system 25 to control the reactor An example of a characteristic value 29 of a characteristic barrier is shown in Fig. 1, which is shown in the form of an area formed by the separated calibration point 30, wherein the amount of the catalyst 4 delivered to the reaction system 5 [by mass percentage) (m_% ΚΑτ) indicates] the temperature A measured by the reactor system 5 and the Coriolis force ρ c measured by the Korotk force measurement system 19. Characteristics of this form of the characteristic field The value 29 can be calculated via the estimation system 27 for each Korotk force measurement system (19 to 22) and the respective assigned reactor system (5, 7 - 10 ' 13 ' 15). For the reactor system An alternative or a supplement to such a single control can produce a multidimensional eigenvalue 29, but it cannot be exemplified, this is a hypersurface, 柯 and the Coriolis force measured by all measurement systems and some or all of the reactors The reaction parameters set by the system are related. With this form of control, the reaction parameters of different reactor systems are combined into a single parameter field, which is simultaneously controlled by the state variable. The separation control of the individual reactors, the advantage of this form of control is also based on the reaction parameters of different reactor systems. Examples of reaction parameters that can be controlled by this method are the first in the reaction system 5. Monomer, first monomer and contact deletion, and pressure in each reactor system: 16 1259839 Residence time and temperature. The following is based on the facility according to Figure 1. Based on polycarbonate, the generation of characteristic value 29 is described by way of example: The optimum value obtained according to the test facility of Figure 1 is added in Table 1, firstly without the use of Korotk force measurement and maintaining stable continuous operation, reaction parameters and The characteristic values of the product are stored in the control system. The Kodak force measurement system is then turned on and the reaction conditions of the facility are kept fixed, and then the reaction conditions in the reactor system are specifically changed in a computationally and well-planned manner to track their The Coriolis force measures the effects and stores their effects. In this method, each Coriolis force measurement system uses the same steps until the characteristic values are calculated in the many state parameters expected during the operation of facility i. The benzoate (PC) is produced from bisphenol A (BpA) as the first monomer and diphenyl carbonate (Dpc) as the first precursor and sodium phenolate as a catalyst. In the operation of facility i, all of the reactor systems have a molar ratio of the two monomers between 〇.9 and 2.Q in the fixed re-injection, pressure and fixed catalyst concentration. Make a gradual change, in each step t change the holding material is set to about two small molar phase: Ke I 都要 in the transition phase or the holding phase must be measured on the Korotkton sensor and The reaction parameters attributed to the reaction system. An initial correction point 30 of the characteristic value 29 measured in this method, of which. Figure 2, 胄, the reaction parameters of the reactor and the reactor system are interrelated Moer ratio, ±... How the Coriolis force measured by the Coriolis force sensor corresponds to, the measured Coriolis force can be found Ear ratio. 17 1259839 & After the ear ratio, the remaining reaction parameters are changed under (), for example, there are ten steps in each case. The temperature of the sputum reactor can be varied between 200 吒 and 35 〇〇c and the pressure varies between 2ba ""0 mbar, and can be obtained under steady state: the output of the value of 50 〇 / 以及 and - 5〇% to +2〇〇% of the catalyst concentration, during the change of the reaction parameters, the remaining reaction parameters remain at the stable operation = value. From these set values, a large number of calibration points can be obtained, and features are also obtained. A value of 29, wherein the reaction parameters are correlated with the amount obtained by the Coriolis force sensor. In the operation of such a facility, the feature value 29 can be used as a multi-dimensional lookup table using a set of measured ke For the query, the change of the reaction parameters must be set in order to achieve the desired stability. In addition, the 'eigenvalue can also be used to train the neural network, and the measured Coriolis force is taken as the input vector and the response is set. The change of the parameter to the set state is taken as the output vector. The function of this kind of neural network is like RiUer and
Helge 等人在"Neuronale Netze: Eine Einftihrung in Neuroinformatik selbstorganisierter Netzwerke1’（類神經網 路：自組織網路的類神經計算學的導論），Addison-Wesley (1991年）中所描述的。控制單元25使用由訓練所產生的類 神經網路來控制設施1。 使用此方法及裝置可以有效地生產聚合物並處理聚合 物生產的次要產物。 以三個測試實施例作為基礎於下面說明依據本發明使 18 1259839 用柯氏力來控制的效果’表1 k供測試實施例的綜覽。 實施例1 實施例1中依據上述的生產方法所得到的最終產物i i 是 PET。 由表1可看出使用柯氏力控制可使產物性質變化的幅 度減小達到20%。 實施例2 依據圖1的測試設施中所生產的最終產物丨丨是聚丁稀 對苯二甲酸酯（PBT)，它是由當作第一單體的對苯二甲酸 (TPA)與當作第二單體的丨，4· 丁二醇（BD)(莫耳比為1:25) 以及100 ppm的鈦觸媒所生產的，質量產量為13〇 kg/h， 一般的反應條件是在220。(：到260〇C之間的溫度以及〇.5 mbar到900 mbar之間的壓力。 由於本發明的控制方法，產物性質變化的幅度可以被 減小達20%以上。 貫施例3 依據圖2的測試設施中所生產的最終產物u是聚碳酸 酯（pc)，它是由當作第一單體的雙酚A(BPA)與當作第二單 體的一苯基碳酸酯（DPC)(莫耳比為1:1·2)以及當作觸媒的 100 PPm的酚化鈉所生產，質量產量為15〇kg/h，反應條件 是在250°C到3 30°C之間的溫度以及〇 3 mbar到950 mbar 19 1259839 之間的壓力。可由表1看出使用導入柯氏力測量的控制可 以提高黃度係數。 表 1 實 施 例 產物 單體 方目. 座里 kg/h (不 使用柯 氏力控 制） 末端基 meq/kg (不使用 柯氏力控 制） 次要產物 (不使用柯 氏力控制） 極限 黏度 g/dl 曰- 座里 kg/h (使 用柯氏 力控制） 末端基 meq/kg (使用柯 氏力控 制） 次要產物 (使用柯 氏力控 制） 1 PET TPA 100 20 土 4 DEG: 0.64 105 17 土 2 DEG: EG 0.4〇/〇 士 0.2 0.2% ±0.1 2 PBT TPA 130 32 土 5 THF: 1.1 140 24 士 2 THF: BDO 25% ± 3 19% 士 1 3 PC ΒΡΑ 150 酚： 黃度係數： 0.58 160 28 土 2 黃度係數： DPC 20 士 6 96.0 ± 2 98.0 土 1 圖式簡單說明 圖1以示意圖顯示一種用於聚合物生產的第一設施， 它使用本發明方法及本發明的測量系統的一種具體實例。 圖2顯示一種柯氏力與單體莫耳比之間的關係的示意 圖。 主要元件之符號說明 1.·設施；2..第一單體；3··第二單體；4··觸媒； 5、7-10、13、15..反應器系統；6.·管件系統；11··產物； 12.. 排放管；14..收集管線；16..裂解產物；17..排放管線； 18·.回流管；19-23··柯氏力測量系統；24··數據路線； 25.. 控制系統；26·.控制路線；27..估算系統；29·.特徵值； 3 0..校正點 20Helge et al., "Neuronale Netze: Eine Einftihrung in Neuroinformatik selbstorganisierter Netzwerke1' (Introduction to Neural Network Computing: An Introduction to Neurological Computing in Self-Organizing Networks), Addison-Wesley (1991). The control unit 25 uses the neural network of the type generated by the training to control the facility 1. Using this method and apparatus, the polymer can be efficiently produced and the secondary products produced by the polymer can be processed. Based on the three test examples, an overview of the effect of the control of the 18 1259839 controlled by Korotk force in accordance with the present invention will be described below. Example 1 The final product i i obtained in Example 1 according to the above production method was PET. It can be seen from Table 1 that the use of Korotk force control can reduce the amplitude of product property changes by up to 20%. Example 2 The final product produced in the test facility according to Figure 1 is polybutylene terephthalate (PBT), which is derived from terephthalic acid (TPA) as the first monomer. As a second monomer, 4, butanediol (BD) (molar ratio 1:25) and 100 ppm titanium catalyst, the mass yield is 13 〇kg / h, the general reaction conditions are At 220. (: a temperature between 260 ° C and a pressure between 5 5 mbar and 900 mbar. Due to the control method of the present invention, the magnitude of the change in product properties can be reduced by more than 20%. The final product u produced in the test facility of 2 is polycarbonate (pc) which is composed of bisphenol A (BPA) as the first monomer and monophenyl carbonate as the second monomer (DPC). (Morby is 1:1·2) and 100 PPm of sodium phenolate as a catalyst, the mass yield is 15〇kg/h, and the reaction conditions are between 250°C and 3 30°C. The temperature and the pressure between 〇3 mbar and 950 mbar 19 1259839. It can be seen from Table 1 that the yellowness coefficient can be increased by the control of the introduction of the Coriolis force measurement. Table 1 Example product monomer order. (Do not use Coriolis force control) End base meq/kg (without Coriolis force control) Secondary product (without Coriolis force control) Ultimate viscosity g/dl 曰- Seat kg/h (Use Coriolis force control End group meq/kg (using Kodak force control) Secondary product (using Kodak force control) 1 PET TPA 100 20 Soil 4 DEG: 0.64 105 1 7 Soil 2 DEG: EG 0.4〇/〇士0.2 0.2% ±0.1 2 PBT TPA 130 32 Soil 5 THF: 1.1 140 24 ± 2 THF: BDO 25% ± 3 19% ± 1 3 PC ΒΡΑ 150 Phenol: Yellowness factor : 0.58 160 28 Soil 2 Yellowness factor: DPC 20 ± 6 96.0 ± 2 98.0 Soil 1 BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows in schematic view a first facility for polymer production using the method of the invention and the measurement of the invention A specific example of the system. Figure 2 shows a schematic diagram of the relationship between the Coriolis force and the monomer molar ratio. Symbols of the main components 1. Installation; 2. First monomer; 3 · Second Body; 4··catalyst; 5, 7-10, 13, 15: reactor system; 6. pipe fitting system; 11·product; 12.. discharge pipe; 14. collection pipeline; Product; 17.. Discharge line; 18·. Return line; 19-23··Coriolis force measurement system; 24··Data route; 25.. Control system; 26·. Control route; 27.. Estimation system; ·.Characteristic value; 3 0..Correction point 20