EP0947625A1 - Process and apparatus for controlling and optimizing the process of chemical recovery during cellulose production - Google Patents

Abstract

A digestion chemicals recovery process for a sulfite or sulphate pulp mill a) measures at least one continuous electromagnetic radiation spectrum (100nm - 400micron), for the process, b) evaluates indices PC1...PCn for the principle components of the liquor, and c) sets up a control model using the calculated values PC1...PCn , plus laboratory / plant measurements. A digestion chemicals recovery process for a sulfite or sulphate pulp mill uses a process control logic system or a control model. The continuous electromagnetic radiation spectrum is measured by either absorption, emission, or Raman spectroscopy. The laboratory / plant measurements comprise: Liquor temperature, concentration, pH, conductivity and flow rate. Preferably data from a number of spectrometers sampling at different stages of the wet process is analyzed e.g. via Fourier and Kubelka Munk techniques. The control system is able to assess the validity of individual measurements before inputting them to the control model; it may use neural networks or fuzzy logic. In the Sulphate Process, the recovery system uses concentration data for active alkali, total sulfide, sulphate, sodium sulfide, sodium carbonate and sodium hydroxide. The Sulfite Process, recovery requires concentration data for sulfur dioxide, bisulfite, sulfite, thiosulfate, sulfate, dead-burnt magnesia and active magnesium oxide. A CLAIM for the equipment required to operate the process includes: At least one spectrometer, a computer for evaluation of PC1...PCn from the continuous spectra and for inserting reference variables PC1...PCn , and optionally the discrete physical and / or chemical properties as process parameters into the process control logic or the control model.

Description

The invention relates to a process control method and for process optimization of chemical recovery using in the manufacture of pulp at least one state model and / or process model. In addition, the invention relates to a device for Execution of the procedure.

The production of pulp by so-called pulping is done by boiling wood chips using appropriate cooking chemicals either in in a continuous or in a discontinuous process. The cellulose boiling with sulfate or sulfite exclusion usually a recovery process for the Cooking chemicals.

The process control in the recovery of the cooking chemicals is difficult because important quality parameters of the cooking chemicals - for example, in the case of sulfate digestion, the active alkali concentration, sulfidity, sulfate content, sulfur content, Na ₂ S, Na ₂ SO ₄ , Na ₂ CO ₃ , Na ₂ O, CaO, NaOH, NaCl, K ₂ CO ₃ , CaCO ₃ , Ca (OH) ₂ and for sulfite digestion, for example, the chemicals SO ₂ , HSO ₃ ^- , SO ₃ ^- , S ₂ O ₃ ^- , SO ₄ ^- , MgO , Active MgO, burnt MgO - only be measured in the laboratory with a time delay. For this reason, an interim faulty production cannot be excluded. In order to avoid the risk of incorrect production, the production process is therefore usually operated with a greater margin of safety with regard to the quality of the cooking chemicals than would actually be necessary.

• Der Prozeß wird nach Erfahrungswerten gesteuert. Regeleingriffe sind nur in geringem Umfang möglich. Die Qualität der Kochchemikalien kann nur im nachhinein bestimmt werden. Eine große Streuung der Qualitätsparameter ist daher häufig nicht vermeidbar.
• Es erfolgt ein analytisches Messen einzelner Komponenten der Kochflüssigkeit mit beispielsweise Titrationsautomaten.

The following strategies are often used in the prior art:

• The process is controlled based on experience. Regular interventions are only possible to a limited extent. The quality of the cooking chemicals can only be determined after the fact. A wide spread of quality parameters is therefore often unavoidable.
• There is an analytical measurement of individual components of the cooking liquid using, for example, automatic titration devices.

Holz", wie Holzhackschnitzel oder Holzfasern, oder direkt am Rohstoff Altpapier".In addition, it has already been proposed to provide infrared spectroscopy for process assessment, for which a number of publications exist. DE 195 10 009 A1 discloses a method for the process control of a paper machine. Selected spectral parameters are used for the measurement of product quality parameters. Neural networks that have been trained on the product parameters measured in the laboratory are used specifically for process control in the headbox of a paper machine. Furthermore, DE 195 10 008 A1 discloses a process control process in pulp and paper production in which spectral parameters of different wavelengths are derived for determining the starting materials and for evaluating the quality of the raw materials. Correction signals for the system-related regulating or control devices are also to be determined, in particular with neural networks. With this method, the measurement signal is recorded directly on the raw material

Wood ", such as wood chips or wood fibers, or directly on the raw material Waste paper ".

According to US 5 378 320 A, sulfate pulp production via IR spectroscopy at selected wavenumbers the absorption is measured and are different via correlations Alkali concentrations determined. It is proposed the lignin content with spectroscopic measurements of pulp to determine.

Furthermore, DE 36 16 051 A1 discloses a method for controlling the digestion of lignocelluloses using FTIR spectroscopy. By measuring in the MIR (M ittleren I nf r arot) range of Restligninanteil and the concentration of the cooking chemicals is determined, for example, the content of free SO ₂ in the sulphite cooking.

In addition, an incinerator is from DE 42 21 404 A1 for the waste liquor of a pulp cooker with a control device known for the combustion air at which the amount at the temperature of a waste liquor atomized in the system Dependency on a measure of the ash quality like this takes place that the combustion products of the waste liquor as far as possible large amounts of hydratable starting materials for the Recovery of digestion chemicals included.

Finally, in US 5 616 214 A, US 5 364 502 A and US 5,282,931 proposed spectroscopy specifically in Connection with the recovery of the digestion chemicals to use in the manufacture of pulp. With these publications are samples from the process taken and examined at defined wavelengths. From that apart from that is proposed with JP 5-321183 A for regulation the combustion of black liquor in a recovery boiler to provide a fuzzy control in such a way that Fuzzy rules drawn up from operational practice and with one appropriate membership function are correlated.

In the unpublished prior art according to DE 196 53 530 A1, DE 196 53 532 A1, DE 196 53 479 A1 and DE 196 53 477 A1 is on the application and evaluation the spectroscopy in the pulp boiling, in the bleaching of Paper materials as well as in the operation of refiners and paper machines received. For these applications, respectively suggested continuous in at least one place Spectra of electromagnetic radiation and / or continuous Measure spectra of mechanical properties still discrete physical in at least one place and / or capture chemical properties through mathematical Evaluation of the continuous spectra parameters to form and from the parameters, the physical and / or chemical properties and laboratory measurements of product properties to set up a state model and possibly a process model.

In contrast, it is an object of the invention, the process control especially for chemical recovery in manufacturing of pulp to improve and an associated device to accomplish.

The task is in a method of the aforementioned Art according to the invention with the features of claim 1 solved, with advantageous developments in the dependent Claims are specified. The associated device is in single property claim specified.

The method with acquisition of the continuous spectra can alternatively be used for the sulfite or sulfate process. In the invention, spectroscopic measurements are carried out specifically on the chemical streams of the recovery system, ie in the sulfate digestion, for example on the soda melt after the recovery kettle, in the green, white and black liquor, and in the sulfite digestion, for example in the ash stream or on the ash or in the cooking liquids, carried out. For the measurement of continuous spectra of electromagnetic radiation, absorption, emission and / or luminescence spectroscopy in the spectral range from 0.1 μm to 400 μm, preferably between 0.4 μm and 100 μm, can advantageously be used. The continuous spectra are evaluated with suitable calculation methods and allow a statement about the expected chemical composition of the cooking liquid. On the basis of these statements, the process can be intervened to regulate, for example by changing the air distribution in the recovery boiler, the CaO supply in the so-called caustifier, by adding so-called Make-up chemicals and by regulating the mixing ratio of black liquor, white liquor and dilution water.

The improved chemical recovery is special advantageous for pulp production with a smoothing the quality of the cooking liquid on high Level. This will increase the yield and production volume, a reduction in chemicals and / or Use of energy as well as the saving of auxiliary materials in the bleaching stage following the pulp production reached.

The invention is compared to the state stated at the outset the technology in essential points, such as in particular Measurement site, the measurement method, the processing of the spectra and the Modeling, further developed.

If electromagnetic waves in the range between 100 nm and 400 µm, preferably in the range from 0.4 µm to 100 µm, are used, so-called Raman spectra can also be measured in addition to absorption, emission or luminescence spectra. For example, the absorption spectroscopy can be carried out diffuse reflection or attenuated total reflection (ATR = a ttennuated t otal r eflection) in transmission. The excitation for luminescence can take place, for example, through the irradiation of electromagnetic radiation, for example UV radiation, or through a specific chemical reaction such as chemiluminescence; on the other hand, the emission is excited, for example, by irradiation with electrons. When measuring in the infrared range (IR: 800 nm to 20 µm), Fourier transform infrared spectroscopy (FTIR) can preferably be used. In the case of inhomogeneous samples, the spectroscopic measurement is carried out several times to compare the plausibility.

• durch Fourier-Transformation.
• Bei der Messung der Absorption durch diffuse Reflexion durch Umrechnung in sog. Kubelka-Munk-Einheiten und Korrektur von Mehrfachstreueffekten
• durch Normierung und Glättung der Spektren
• durch Ermittlung von für die Modellbildung ungeeigneten Spektren. Die Ausschaltung ungeeigneter Messungen kann z.B. durch Vergleich mit Referenzspektren erfolgen
• durch Bildung von Mittelwerten bei mehreren Spektren zu einer Probennahme.

The spectra are preferably preprocessed as follows:

• through Fourier transform.
• When measuring the absorption by diffuse reflection by converting it into so-called Kubelka-Munk units and correcting multiple scattering effects
• by normalizing and smoothing the spectra
• by determining spectra that are unsuitable for modeling. Unsuitable measurements can be switched off, for example, by comparison with reference spectra
• by averaging multiple spectra for one sample.

• PCA-Verfahren (principle component analysis) oder auch sog. Hauptkomponentenanalyse
• PLS-Verfahren(partial least square), einer der Fachwelt bekannten Rechenmethode unter Verwendung kleinster Quadrate
• Neuronale Netze
• Analytische Beschreibung der Spektren, z.B. im Bereich des Infraroten (IR) durch Lage, Intensität und Breite der wichtigsten Absorptions- oder Emissionspeaks, Ermittlung dieser Größen z.B. mit einfachen Minimum-/Maximum-Verfahren oder der
• zweite Ableitung der Spektren

Die Kenngrößen werden zur Modellierung der gewünschten Qualitätsparameter herangezogen.After these first processing steps, the following computer-aided methods for determining parameters can be applied to the spectra in whole or in sections, the description of the spectra essentially being carried out by their main components:

• PCA-method (p rinciple c omponent a nalysis) or so-called. Principal Component Analysis
• PLS method ( p artial l east s quare), a calculation method known to the experts using the least squares
• Neural Networks
• Analytical description of the spectra, for example in the infrared (IR) range by position, intensity and width of the most important absorption or emission peaks, determination of these quantities, for example using simple minimum / maximum methods or
• second derivative of the spectra

The parameters are used to model the desired quality parameters.

The state models for calculating the quality parameters can with a sufficiently large number of data based of neural networks, fuzzy systems, multilinear regression models or combinations thereof. Alternative to purely data-driven models are also Combined models possible, where additional analytical Knowledge is brought in. In the same way and with The process models can also be constructed using the same means become.

Laboratory measurements of intermediate and end products are used to set up the models. Training the models respectively their validation is based on of laboratory values and can be in certain time intervals can be repeated, with only partial re-learning is possible.

A model validity check integrated in the spectrum preprocessing ( Novelty detection ) according to the older, unpublished DE 196 322 45 A1 can indicate the need for a new training phase in good time in the ongoing process.

Figur 1

ein Schema der Meßstellen bei der Chemikalienrückgewinnung,

Figur 2

als Ausschnitt von Figur 1 ein entsprechendes Schema für die Kaustifizierung,

Figur 3

ein kontinuierliches Spektrum von optischen Messungen an Kochflüssigkeit,

Figuren 4 und 5

jeweils ein Zustandsmodell für die Qualitätsparameter der Kochflüssigkeit für den Sulfataufschluß und den Sulfitaufschluß,

Figur 6

ein Prozeßmodell für die Chemikalienaufbereitung beim Sulfatzellstoff,

Figur 7

ein Prozeßmodell für die Chemikalienaufbereitung beim Sulfitzellstoff,

Figur 8

ein dynamisches Prozeßmodell,

Figur 9

den schematischen Aufbau einer Prozeßoptimierung zum Steuern der Chemikalienaufbereitung,

Figur 10

eine Variante zu Figur 8 unter Einbeziehung des dynamischen Modells gemäß Figur 7,

Figur 11

schematisch die Vorverarbeitung und Verdichtung der Spektren und

Figur 12

eine Vorrichtung zur optimierten Prozeßführung bei der Chemikalienaufbereitung.

Further details and advantages of the invention result from the following description of the figures of exemplary embodiments with reference to the drawing. Show it

Figure 1

a diagram of the measuring points in chemical recovery,

Figure 2

1 shows a corresponding scheme for caustification,

Figure 3

a continuous spectrum of optical measurements on cooking liquid,

Figures 4 and 5

each a state model for the quality parameters of the cooking liquid for the sulfate digestion and the sulfite digestion,

Figure 6

a process model for the chemical processing of sulfate pulp,

Figure 7

a process model for the chemical processing of sulfite pulp,

Figure 8

a dynamic process model,

Figure 9

the schematic structure of a process optimization for controlling the chemical preparation,

Figure 10

8 shows a variant of FIG. 8, including the dynamic model according to FIG. 7,

Figure 11

schematically the preprocessing and compression of the spectra and

Figure 12

a device for optimized process control in chemical processing.

The figures are partially described below together. The same or equivalent parts have corresponding Reference numerals.

In Figure 1, the structure of a continuously working Chemical preparation shown, in the present Connection the measuring point scheme is clarified. Here is used by a well-known stove for the production of pulp by cooking raw materials, especially in the form of Wood chips, in a suitable cooking liquid. At 11 is the inlet for the wood in the stove 10, with 12 the feed line for the cooking liquid and with 13 designated the outlet for the finished pulp. The cooker 10 close a blow tank 14 and several washers 15, 15 ', 15' 'with assigned filtrate tanks 16, 16', 16 ''. From The finished pulp is output to the washer.

The filtrate is passed through at least one unit 17, advantageously units 17, 17 ', 17'', ... for multi-stage evaporation, into a so-called black liquor tank 18 and from there into a recovery tank 20 for cooking chemicals. In the recovery boiler 20, the non-recyclable waste liquors, referred to as black liquor, are to be burned and the recyclable cooking chemicals are to be recovered. The fly ash is removed via an electric filter 21. The soda melt accumulating at the bottom of the recovery boiler 20 flows into a melt-dissolving tank 22. This is followed by a green liquor clarifying tank 23 and a causticizing system in which chemical conversion reactions for the extraction of caustic take place. This system essentially consists of an extinguishing tank 24, several individual caustifiers 25, 25 ', 25'', a white liquor clear tank 26 and a cooking liquor tank 27. The system for caustification is assigned a scrubber 28 with filter 29 and a lime kiln 30. The lime burned there is returned to the extinguishing tank 24. The mode of operation of the staggered caustifiers 25, 25 ', 25''and25''' is clear from the partial illustration according to FIG. There is a computer 105 for recording and processing characteristic data, in particular also physical or chemical properties, such as flow m ° and density ρ.

Due to the intended continuous operation of the The chemical preparation shown in the system is comparative complex: it can alternatively follow the sulfate process or be worked after the sulfite process. Here is at various significant points in the process chain to grasp the process status and the process flow optimize what must be measured values. In Figure 1 are spectrometers at appropriate points in the process chain for recording individual continuous spectra A, B, C, D, E, F, G, H and I of electromagnetic radiation and in the partial representation according to FIG. 2 with corresponding ones Modifications spectrometer at suitable measuring points for the Spectra F 'and G' attached. With this arrangement, preferably in the infrared range (IR: 1 - 25 µm), Spectroscopic measurements online in the cooking liquid carried out. Such a measurement can e.g. with the help of a known ATR probe or a known FTIR spectrometer be performed.

The measured spectra are processed by signal preprocessing smoothed and standardized. Then a disassembly in the main components and / or the identification more important Peaks occur. From the main components you can then with models, e.g. using methods of multilinear regression, the concentrations of the following quantities are calculated: Effective alkali, sulfidity, carbonate, sulfate and thiosulfate.

As is known, the latter values can be found in Process models for the calculation of the quality parameters of the Cooking liquid 12 produced in Figure 1 can be used.

A regulatory intervention is at various points in the Recovery process conceivable, e.g. by replenishing Chemicals, by temperature variations or by Changes in the Circulations.

In FIG. 3, 31 characterizes the course of an IR spectrum of cooking liquid in the wavenumber range 1500 to 900 cm ⁻¹ . To obtain spectra A to I in FIGS. 1 and 2, measurements can also be carried out in other wavenumber ranges.

The spectra shown by way of example in FIG. 3 using a wide variety of mathematical methods, each of which is known from the prior art, processed As already mentioned, this is essentially the case a preprocessing of the spectra, the introduction of analytical knowledge and possible outlier detection, for a correct intended determination of parameters. It is essential that from the continuous spectra alone by a suitable mathematical evaluation of such parameters especially for the chemical flows in the recovery process can be formed and that from the parameters and laboratory measurements the chemical concentrations a state model and / or additionally a process model with process properties is formed. Furthermore, this can be done on the chemical flows discrete physical and / or chemical properties be recorded for further processing.

The latter applies to cooking chemicals both in the sulfate process and in the sulfite process: In the sulfate process, the chemical concentrations describing the chemical recovery process are in particular the active alkali concentration, the sulfidity, the sulfate content and the content of Na ₂ S, Na ₂ CO ₃ and NaOH used. When sulfite are used as the chemical recovery process described chemical concentrations of the acids, the total SO _2, HSO ₃ ^- -, -SO ₃ ^- -, S ₂ O ₃ ^- -, SO ₄ ^- concentrations and for the bases, the MgO Concentration, divided into so-called burnt MgO and active MgO.

In FIGS. 4 and 5, 40A and 40B each represent a relevant state model of the cooking chemicals, from which their properties, such as the concentration, are calculated. The essential quality parameters for the cooking liquid to be prepared can be derived from this. For this purpose, are determined from the spectra with either the so-called principal component analysis (PCA = p rincipal c omponent a nalysis) or with the so-called PLS method (p artial l east s quare) the characteristics and input to the model.

4 and 5 are selected discrete ones mechanical or chemical properties and in the state model 40 entered. The state model 40A or 40B can but possibly alone from the continuous spectra determined parameters PC1 to PCn are formed.

The model 40A or 40B can e.g. advantageously by a neural network should be formed. When modeling overall of modern information technology computing methods, such as especially evolutionary or genetic Algorithms, made use of.

In Figure 6 is a process model for chemical recovery designated by the sulfate process at 50. Entered are discrete physical and chemical properties, e.g. Temperature, pH and Pressure and the state variables of the cooking liquid as an output of the state models, for example according to FIGS. 4 or 5. The process models can also be used only with those from the Characteristics determined continuous spectra formed become.

A corresponding process model 60 is special in FIG specified for the sulfite process. Unless discrete physical and / or chemical properties are used in addition to the optical spectrum, for example the flows and / or temperatures are given, which is shown in Figure 6. The pH value during sulfite digestion can be entered.

For each quality parameter to be calculated are expedient create your own models. To increase the prediction accuracy can sub-models for each parameter be formed.

According to FIG. 8, the process model according to FIG. 6 can also be used as dynamic process model 70 can be designed. The input variables are here according to the input variables in FIG. 6 in each case at discrete times k, ..., (k-n). The same applies to the chemical concentration. Out of it Dynamic process model 70 determines the chemical concentration at future times (k + 1.

When setting up the models can be of neural Networks can also be used by fuzzy methods. It combined neuro-fuzzy systems can be used.

After the state and / or process models have been drawn up, it is particularly important to validate the validity of the models, which can be done by online training of the individual models or of the partial models. In practice, it can be important to check the results obtained by computer-aided selection of all information-bearing data. This process was called Novelty Detection "suggests and enables new data records to be incorporated into the evaluation process during the ongoing production process. If the results are not consistent, the models have to be retrained.

The obtained using the described modeling procedures Sizes are used for process control and process optimization in the recovery plant. For this is the in Figure 9 Process generally designated 80, which is the current Process state based on the spectra with 81 results. The process model is denoted here by 82, from which the data is converted into a Unit to be given to cost function 83, the same time with data for costs and prices from unit 84 is applied. An optimizer 85 determines the Actuating variables 86, which are fed back into the process model 82 and continue to be the optimal manipulated variables 87 for process control. These can also be switched via a switch 88 Plant operators can be switched through.

The corresponding results from FIG. 10 for process control, at which equally a dynamic model accordingly of Figure 8 is used. Here is an additional one Unit 89 with the dynamic model in which the current process status entered on the one hand and the optimal On the other hand, manipulated variables are specified.

With reference to FIG. 11 it is clarified that a unit 91 for Pre-processing and compression for the spectra of the entire spectrum 90 serves, from which e.g. the parameters PC 1 to PC 10 are calculated. In the The main component analysis used for this is from a suitable number of spectra, for example between three and ten spectra, so-called scores, for the purpose of data reduction educated. From this, the input variables PC1, ..., PCn in particular of the model calculated according to FIG. 4.

According to FIG. 11, the parameters PC1 to PCn flow into the state model 93 and in the process model 94, whereby here additionally discrete physical and / or chemical properties and the process description the state model for Complement the process model. From the state model 93, the product properties the cooking chemicals and from the process model 94 derived the product properties of the cooking liquid. Now all chemical additives can be calculated and in Optimized to achieve the desired cost savings.

Figure 12 shows how with appropriate evaluation and optimization software using a computer that the computer 105 from Figure 1 can be intervened in an existing process control system becomes. Optimized manipulated variables can be generated a known automation device 100 as a process control system charge that with the actual system for Implementation of the process interacts. Basically So the existing chemical recovery facilities through several spectrometers 101 to 103 and an associated, optimization program implemented as software, that runs on the computer 105, added.

Claims (28)

Process control and process optimization process for chemical recovery in the manufacture of pulp using at least one status model and / or process model, with the following features: a) continuous spectra of electromagnetic radiation are measured at at least one point on the chemical flows, b) through mathematical evaluation of the continuous spectra, parameters (PCl ... PCn) for the chemical flows are formed, d) the status model is established from the parameters (PCl ... PCn) and laboratory measurements of the chemical concentrations and / or the process model is additionally established with process properties. Method according to Claim 1, characterized in that discrete physical and / or chemical properties are recorded at at least one point on the chemical streams. Method according to Claim 2, characterized in that the discrete physical and / or chemical properties are used to set up the state model and, if appropriate, the process model. Method according to claim 1, characterized in that measurements are made at wavelengths of electromagnetic radiation between 100 nm and 400 µm. A method according to claim 4, characterized in that the electromagnetic radiation is detected as an absorption, emission, luminescence or Raman spectrum. A method according to claim 4, characterized in that the electromagnetic radiation is detected in transmission, direct or diffuse reflection or attenuated total reflection (ATR). Method according to Claim 2, characterized in that the electrical conductivity, the pH value, the temperature, the flow rates and the chemical concentrations of the chemical flows are recorded as discrete physical and / or chemical properties. Method according to Claim 1, characterized in that a main component analysis is carried out on a predetermined number of spectra and a corresponding number of scores is selected for data reduction, and that the parameters for the model formation are determined therefrom. Method according to Claim 8, characterized in that the spectra are preprocessed and compressed, and in that the specific characteristic values of the spectra, in particular the main components, are selected for describing the product state and are entered directly into the state model. Method according to Claim 8, characterized in that the spectra are preprocessed and compressed, that the specific parameters of the spectra, in particular the main components, are introduced into the state model, and that the product properties are formed at the output of the state model and are entered directly into the process model. Method according to Claim 1, characterized in that spectra which are unsuitable for the model formation are eliminated by a plausibility check. Method according to one of the preceding claims, characterized in that neural networks and / or fuzzy logic are used for modeling. Method according to claim 1, characterized in that the measurements of the spectra in the sulfate process are carried out both on the soda melt and on the aqueous chemical streams such as green liquor, white liquor and black liquor. Method according to Claim 1, characterized in that the spectra are measured in the sulfite process both on the ash flow or on the ash and on the cooking acids. A method according to claim 13 or claim 14, characterized in that the parameters obtained by evaluating the spectra are used to control and / or regulate the recovery process. A method according to claim 15, characterized in that the quality parameters of the finished cooking liquid, in particular the chemical concentrations, are modeled to control and / or regulate the chemical recovery. A method according to claim 16, characterized in that the model approaches are used not only to predict the product quality but also to calculate the chemical inputs. A method according to claim 16, characterized in that the chemical alkali concentration, the sulfidity, the sulfate content and the content of Na ₂ S, Na ₂ CO ₃ and NaOH are used as chemical concentrations describing the chemical recovery process in the sulfate process. A method according to claim 16, characterized in that as chemical concentrations describing the chemical recovery process in the sulfite process for the acids the total SO ₂ , the HSO ₃ ^- -, SO ₃ ^- -, S ₂ O ₃ ^- -, SO ₄ ^- - Concentrations and for the bases the MgO concentration, divided into fired MgO and active MgO, can be used. A method according to claim 16, characterized in that the model approach is used with the quality parameters in the process optimization. Method according to one of the preceding claims, characterized in that a cost function is formed which is optimized with an optimizer by suitable variation of the manipulated variables. A method according to claim 19, characterized in that the optimization is carried out by genetic algorithms. Method according to Claim 20, characterized in that a cost function for the production costs and / or a profit function is used as the cost function. Method according to one of the preceding claims, characterized in that a dynamic model is used to check the manipulated variables optimized by a static model, a neural network in particular being used as the dynamic model. Method according to one of the preceding claims, characterized in that the model and / or the sub-models are trained online. Method according to one of the preceding claims, characterized in that the results obtained are checked by computer-assisted selection of all information-carrying data (

Novelty detection ). Method according to Claim 26, characterized in that retraining is carried out if there are inconsistent results. Device for performing the method according to a of claims 1 to 27, consisting of at least one Spectrometer (101, 102, 103), from a digital computer (105) for the mathematical evaluation of the continuous spectra for calculating the parameters (PC1 ... PCn) and for installation of the state model and / or the process model the parameters (PC1 ... PCn) and possibly the discrete ones physical and / or chemical properties as process properties, and from a process control system (100).

Images