阅读说明：本技术 用于操作连续生产线的方法 (Method for operating a continuous production line ) 是由 吴海 于 2018-07-11 设计创作，主要内容包括：本发明涉及一种用于操作连续生产线的方法,该方法包括使用计算机辅助的动态性能预测控制模型生产连续加工轧制钢带的退火步骤。(The invention relates to a method for operating a continuous production line, comprising an annealing step for producing a continuously processed rolled steel strip using a computer-assisted dynamic performance prediction control model.)

1. Method for operating a Continuous Production Line (CPL) comprising an annealing step for producing a continuously processed rolled steel strip using a computer-aided dynamic property prediction control model (DPPC) comprising a Material Property Model (MPM) and a Dynamic Processing Model (DPM) predicting a set of Preliminary Processing Settings (PPS) with which target mechanical, surface and geometrical properties for a strip section can be obtained from the material property model, the dynamic processing model determining whether these preliminary processing settings can be achieved in the continuous production line, wherein the output of the material property model is used as an input parameter for the dynamic processing model to enable the dynamic processing model to provide a Final Processing Setting (FPS) for the continuous production line, wherein the input parameters of the material property model include:

one or more elements of a production plan, which is a sequence of strips to be produced, wherein each of said strips possesses its own set of properties, such as strip dimensions, chemical composition, entry conditions and in-strip inhomogeneities resulting from previous processing, said target mechanical properties, surface properties and geometrical properties;

target values of the incoming rolled strip section for the mechanical, surface and/or geometrical properties;

previous processing conditions of the segment and/or one or more parameters of predicted or measured mechanical, surface and/or geometric properties;

(ii) (optionally) one or more elements of the chemical composition of the section of the steel strip;

(ii) optionally one or more microstructure parameters of the section of the incoming rolled strip;

(optional) device condition parameters;

(optionally) feedback parameters from on-line property measurements of the incoming rolled strip;

(optional) feedback parameters from off-line performance measurements, wherein the input parameters of the dynamic process model include:

the output of the material property model; and one or more of:

device condition parameters;

one or more of the input parameters of the material property model;

optionally feedback from on-line continuous line process measurements,

to generate the final processing settings for the steel strip section and wherein the final processing settings for the section are fed into the continuous production line computerized processing automation, whereby the subsequently produced steel strip section meets the target mechanical, surface and/or geometrical properties.

2. The method according to claim 1, wherein the final machining settings of the segment are optimized in an iterative process, wherein the output of the dynamic machining model is fed back into the material property model for the next iteration until the difference between the target mechanical, surface and/or geometrical properties of the segment and the achievable mechanical, surface and/or geometrical properties of the segment has reached a preset minimum value.

3. The method of claim 1 or 2, wherein the incoming rolled strip is a hot rolled strip.

4. The method of claim 1, 2 or 3, wherein the incoming rolled strip is a cold rolled strip.

5. The method of any one of claims 1 to 4, wherein the continuous production line comprises one or more of:

a continuous annealing step;

a first coating processing step;

a temper rolling and/or stretch straightening processing step;

a second coating processing step;

post-processing steps.

6. The method according to claim 5, wherein the first and/or second cladding processing step is a hot dip coating process or an electroplating process.

7. A method according to claims 5-6, wherein the dynamic performance predictive control model also provides automated final tooling settings of the continuous production line for thickness control of the cladding.

8. Method according to any one of claims 5 to 7, wherein the first and/or second coating processing step is an organic coating step.

9. Method according to any of claims 5 to 8, wherein the first and/or second coating processing step is an inorganic coating step, such as Physical Vapour Deposition (PVD).

10. The method of any of claims 1 to 9, wherein the processing settings are recalculated when the production plan or the input parameters of the dynamic performance predictive control model are changed.

11. The method of any of claims 1 to 10, wherein the recalculation of the processing settings is performed at least once every 30 minutes.

12. Computerized processing automation for a continuous production line, wherein the processing automation is implemented such that during operation it performs a method according to one of claims 1 to 11.

13. A continuous production line controlled by a computerized process automation according to claim 12.

14. The continuous production line according to one of claims 1 to 13, in which, during operation, final machining settings of incoming steel strip sections are determined, set and implemented by the continuous production line computerized machining automation, so that the production of the steel strip sections can meet the target mechanical, surface and/or geometrical properties.

Technical Field

The present invention relates to a method for operating a continuous production line.

Background

A Continuous Processing Line (CPL) consists of various processing steps, which typically include annealing, hot dip coating, temper rolling and stretch straightening, in no particular order. These processing steps are used together for a variety of processing purposes to transport the strip product in the intermediate state and its final state with good mechanical, surface and geometrical properties. Failure to meet the intermediate conditions can compromise subsequent processing and related performance. Which can lead to degradation or rejection of the associated end product and cause economic penalties. Therefore, controlling the tooling settings is increasingly important in attempting to control the final properties of the final product. Recently, emphasis has been shifted from process control alone to performance control of the final product. As a result, control models that relate process settings to the resulting performance are becoming increasingly important, and typically rely on the input of sensors that measure one or more process parameters, such as temperature or thickness.

As disclosed in WO2014187886, performance control based on sensor measurements placed at the end of the processing section causes large dead times in the control loop due to the transport of the strip through the production line in long distance channels. This can result in a significant loss of material before the controller detects and corrects the deviation. In some cases, for example, short lengths of strip material have passed through the critical heat treatment temperature regime before the strip head reaches the performance measurement point. Thus, the controller is not active and may take action late.

Model prediction based performance control as disclosed in EP2742158 provides a solution to compensate for dead time in the control loop. The solution uses a performance model to predict performance on-line from process history and adjust the rest of the processes in the production line if necessary to correct quality deviations that would otherwise exist based on the prediction. It does not take into account the upcoming processing dynamics, constraints and disturbances caused by the planned scheduled strip transitions. It is understood from production practice that these operating parameters are the main factors that are realistic and achievable by the process presets determined by the performance models or empirical routines. Therefore, if the upcoming machining situation is not considered, the control actions determined solely by the performance model may be misjudged, infeasible and result in unstable operation. For example, annealing the thick strip a before annealing the thin strip B requires increasing the line speed to compensate for temperature overshoot that may otherwise occur at thickness transition and to minimize the risk of thermal buckling. However, due to limitations in the gas pressure of the wiping process or, in the case of electroplating, due to expected differences in plating thickness, the post-annealing galvanizing process needs to reduce the line speed from the same strip a with a high plating thickness target to a strip B with a low plating thickness. Performance models do not satisfactorily resolve such conflicts.

The raw materials of a continuous production line (for example, the final products of which serve the automobile market or the packaging market) vary in terms of strip size, material chemistry, entry conditions and spatial inhomogeneities in the strip resulting from previous processing, final product properties and corresponding processing requirements. The possible combination of all these factors creates an endless variation of the processing requirements, which vary from order to order and also from strip to strip. The size of the individual order is reduced when the on-time delivery is normal. Thus, processing in a continuous production line is typically transient to a considerable extent and therefore dynamic. The task of mitigating the impact of transient processing conditions on product performance requires skilled operators and requires them to be given constant attention. If the upcoming dynamics of the process are not properly considered or more preferably predicted, the control method or automated system may result in inappropriate control recommendations being given, incorrect control operations, and thus results that are not ideal.

A problem with current process control systems is that the typical setup and feedback model is only sufficient if many materials are produced that are substantially the same. In this case, the variation between subsequent rolls is minimal in terms of mechanical, surface and geometrical properties. However, as batch sizes are reduced, there is a general development that increasingly produces unique tapes that have transient mechanical, surface, and geometric properties even in length. Using typical settings and feedback means that e.g. annealing temperature variations for the strip head will be untimely. Such long dead times can lead to potential product degradation and thus yield loss.

Disclosure of Invention

The object of the present invention is to provide a method for shortening the dead time in the process control of a continuous production line.

It is another object of the invention to provide a method that enables the production of continuously processed end products with an increased yield.

It is another object of the present invention to provide a method that enables the production of continuously processed end products with tighter tolerances on mechanical, surface and/or geometrical properties.

Another object of the invention is to provide a method that is able to address the dynamics of process control in modern production of continuously processed steel.

One or more of these objects are achieved by a method for operating a Continuous Production Line (CPL), comprising an annealing step for producing a continuously processed rolled steel strip using a computer-aided dynamic performance predictive control model (DPPC) comprising a Material Performance Model (MPM) and a Dynamic Processing Model (DPM), wherein the output of the MPM is used as an input parameter of the DPM to enable the DPM to provide a Final Processing Setting (FPS) for the CPL, wherein the input parameters of the MPM comprise:

one or more elements in the production plan;

target values for mechanical, surface and/or geometric properties of the incoming rolled strip section;

previous processing conditions of the section and/or one or more parameters of predicted or measured mechanical, surface and/or geometric properties;

one or more elements of the chemical composition of the segment of (optionally) steel strip;

one or more microstructure parameters of the section of (optionally) incoming rolled strip;

(optional) device condition parameters;

(optionally) feedback parameters from on-line performance measurements of the incoming rolled strip;

(optional) feedback parameters from off-line performance measurements, wherein the input parameters of the DPM include:

the output of the MPM; and one or more of:

device condition parameters;

one or more of the input parameters of the MPM;

optionally feedback from on-line CPL process measurements,

to generate FPSs for the steel strip segments, and wherein the FPSs for the segments are fed into a CPL computerized machining automation, whereby the subsequently produced steel strip segments meet target mechanical, surface and/or geometric properties.

Examples of previous processing conditions may be the crimping temperature of the section or the reduction of the cold rolling mill.

It should be noted that MPM may not require input regarding chemical composition or previous processing conditions or microstructure parameters if the mechanical, surface and/or geometrical properties of the incoming rolled strip segment are accurately predicted or measured. The output of the MPM, in addition to one or more of the device condition parameters, one or more of the input parameters of the MPM (e.g., planned), and (optionally) feedback from on-line CPL process measurements, will be provided to the DPM to predict the FPS with the DPM.

Preferred embodiments are provided in the dependent claims.

The problem of dead time in the control loop is effectively solved with the method according to the invention. It also effectively includes knowledge and prediction of the impending processing dynamics, constraints and disturbances imposed by the planned scheduled strip transitions and knowledge and prediction of the impact of the impending processing dynamics, constraints and disturbances imposed by the planned scheduled strip transitions on product performance. The process control can be optimized by repeatedly examining the performance prediction and the process prediction against the performance objective and the operational objective. By integrating annealing with optional plating, temper rolling, stretch straightening processing, and plating processing, CPL is enabled to produce strip with superior qualities including mechanical, surface, and geometric properties. The method according to the invention is based on the determination of the processing settings for the strip section. The length of the length may vary from strip to strip. A strip may comprise only one or more sections. The number of stages required per strip is determined by DPPC as the case may be or provided by previous processing. For example, DPPC may be used in the same section as in the previous hot rolling process.

The processing plan is the sequence of strips to be produced. Each strip possesses its own set of properties, such as strip dimensions (e.g., width, length, thickness), chemical composition, entry conditions and intra-strip spatial non-uniformities resulting from prior processing (hot, pickling, cold rolling, etc.), target mechanical, surface, and geometric properties. In order to achieve the target mechanical, surface and geometric properties, the processing settings of a continuous production line must be carefully selected while simultaneously taking into account the previous equipment and processing conditions.

Figure 1 shows a schematic view (dashed box) of a continuous production line. The CPL may include a combination of the processing steps shown therein, wherein the strip section travels from left to right. In its most limited embodiment, CPL only includes successive annealing steps. However, such a continuous annealing step is usually followed by a temper rolling step and/or a post-processing step, such as oiling or lubricating. In many cases, the continuously annealed strip may be coated by a hot dip coating process or an electroplating process, and then may be temper rolled and/or coated with an organic or inorganic coating. The flexibility of the method according to the invention makes it possible to customize the DPPC for each CPL.

It is noted that the sequence in fig. 1 is a general sequence of processing steps, but the method according to the invention also allows a different sequence. It is conceivable that a first cladding step is coated with a metal cladding and then temper rolling is carried out, while a second cladding step is coated with an organic or inorganic cladding or a metal cladding, followed by an organic cladding and then temper rolling is carried out.

Fig. 1 shows a forced annealing step and optional subsequent processing steps:

continuous annealing step (mandatory)

First coating process

Temper rolling and/or stretch straightening

Second coating working

Post-processing step

In one embodiment, the first and/or second cladding processing step is a hot dip coating process or an electroplating process.

In one embodiment, the DPPC also provides a CPL automated FPS for thickness control of hot dip coatings.

In an embodiment, the first and/or second coating processing step is an organic coating step.

In one embodiment, the first and/or second coating processing step is an inorganic coating step. In a preferred embodiment, the inorganic coating step comprises or is processed by Physical Vapor Deposition (PVD).

Post processing steps may include processing such as passivation, oiling, etc.

In fig. 2, a Dynamic Performance Predictive Control (DPPC) model is schematically shown. By inputting relevant prior process parameters, target mechanical properties, surface and geometric properties, production plans, microstructures, chemical compositions, etc., a Material Properties Model (MPM) predicts a set of rough process settings (PPS) with which target mechanical, surface and geometric properties for a strip section can be obtained from the MPM. The PPS's (indicated by thick black arrows in fig. 2-4) are provided to a Dynamic Processing Model (DPM) that determines whether the PPS's can be implemented in the CPL. If these PPS's can be implemented in the CPL, they are passed from the DPPC to the control system of the CPL as Final Processing Settings (FPS). The line labeled "a" in fig. 2-4 illustrates that DPM may also use one or more input parameters for MPM in the calculation of FPS.

If the DPM determines that the desired processing settings proposed by the MPM cannot be achieved in the CPL, an iterative process is initiated in which the MPM proposes a new set of processing parameters to the DPM based on what is achievable until an optimal solution is obtained, which is then passed to the control system of the CPL. The best solution is one in which the difference between the target performance and the achievable performance is less than a preset minimum. As long as the preset minimum has not been reached, the iteration as shown by the dashed line in fig. 3 is repeated. If the minimum is not reached at all, the DPPC will suggest introducing a virtual strip between two appropriate strips in the plan, wherein the virtual strip acts as a sacrificial strip, allowing a certain transition to be achieved under the process conditions, while the virtual strip runs through the CPL, so that when the first section of the next strip enters the CPL, the FPS is the best choice to achieve the target properties of that section.

DPM is used to predict upcoming processing conditions. The model describes the conditions and dynamics of the installation and the process, including heat transfer in the annealing process, gas-metal reactions on the strip surface, and, if applicable, hot-dip coating process, wiping process after hot-dip coating, temper rolling and/or stretch straightening.

MPM is used to predict intermediate and final states of strip properties including mechanical properties (grain size, yield strength, tensile strength, elongation, phase fraction, etc.), external/internal oxidation and surface wettability of hot dip coatings and/or geometry based on input processing conditions (surface roughness, shape).

These calculations are performed in real time to determine tooling settings (and tolerances) based on tooling and performance predictions for a given production plan.

It is necessary to optimize the tooling set-up so that the target mechanical, surface and geometric properties of the successive strips are achieved with acceptable consistency and the tooling changes and dynamics required by the tooling set-up can be performed and meet operational targets.

If any of the requirements are violated, an exception or alarm may be raised to notify of the discrepancy. This may result in a rearrangement of the production sequence or a request for virtual or unconstrained strip material if economically advantageous. This may be the case, for example, if a very large difference in annealing temperatures between the strips is required. If not, a virtual tape can be inserted between the two tapes, which then serves as a sacrificial tape.

The present invention uses a process model and a performance model to predict an upcoming production, which includes multiple strips and continuous transitions, given a production plan. This is the only way to be able to predict the progress of the process as close as possible to the actual situation and thereby optimize the process settings and calculate control measures to simultaneously meet multiple process objectives and production goals. By doing so, the hit rate of the production can be maximized to yield a strip product with superior performance.

DPPC with iterations indicated by dashed lines is shown in fig. 4, which also shows other input sources for DPPC, such as off-line product test results, results of on-line processing and product measurements, and device conditions. With these results, both DPM and MPM can be improved.

In a method according to the invention relating to Dynamic Performance Predictive Control (DPPC), the following steps can be distinguished:

a) collecting and receiving input;

b) calculating and predicting preliminary processing settings and, if necessary, intermediate and final properties based on given production plans, target properties, and other inputs;

c) calculating, verifying and adjusting the preliminary processing settings as necessary given the production plan, device conditions and other input conditions;

d) iterating steps b) and c) to optimize the rough machining settings to maximize performance and operational objectives, generating exception notifications when an objective violation is anticipated and unavoidable;

e) calculating a final processing setting for the upcoming strip and transition;

f) process and performance measurements are optionally collected and received, and the DPPC model is altered as necessary to improve the accuracy of the DPPC model, or the accuracy of one of the MPM or DPM.

In the loop operation for the real-time control application, it starts with step a) to collect inputs relating to production plans, raw material (strip) properties and conditions resulting from previous processing, and final product performance requirements and equipment conditions, as well as maintenance, available capacity, etc., and provide them to the MPM and DPM for iterative calculations. The calculation includes steps b), c) and d) to maximize performance and operational objectives. The tooling settings for a given production plan are one of the results of the calculations. The final tooling set-up will be sent to the automated and/or control instrumentation system of the production line. The exception notification will be sent to the production unit, which will generate a production plan to take appropriate action or modify the plan. Process condition measurements and intermediate and final performance measurements (off-line and on-line) are collected periodically to accommodate parameters of MPM and DPM. Preferably, the cycling operation is performed at least once every 30 minutes, preferably in the range of once per minute to once per second.

Conventional process control instrumentation is provided in the CPL to measure process parameters such as line speed and temperature to ensure that FPS is achieved.

The tooling set-up will vary from strip to strip and from section to section in the strip. The calculation of the processing settings for a particular strip or segment of strip is done before loading the strip on the unwinder at the entrance of the production line, preferably before determining the production sequence of the strip. The process settings must be recalculated when a change occurs in the production plan or input, or when the actual process deviation exceeds the tolerance of the newly released process settings due to unforeseen equipment failures, manual intervention in the process speed, etc.

The invention relates to a method for controlling a continuous production line of an annealing furnace for producing a metal strip having excellent properties. The method is run on a computer system comprising:

-data input comprising a production plan comprising a plurality of strips listed in a production sequence, wherein the strip dimensions, material chemistry, entry conditions and in-strip spatial inhomogeneities resulting from previous processing (hot, acid and cold etc.), final product properties and corresponding processing requirements.

-process models describing and predicting the state and dynamics of the apparatus and process including heat transfer during annealing, gas-metal reactions on the strip surface, coating processing (e.g. hot dip galvanization, electroplating), coating thickness adjustment (e.g. zinc wiping process after galvanization or current density), temper rolling and/or stretch straightening.

-a material properties model describing and predicting intermediate and final states of strip properties (including mechanical properties, external/internal oxidation and surface wettability of galvanising or plating, and/or geometry) using the conditions provided by the process model.

-executing an optimization algorithm by iteratively examining the processing predictions and the performance predictions to produce processing settings for a given production plan. The tooling set-up may vary from strip to strip and from segment to segment in the strip. The calculation of the processing settings for a particular strip or strip section is done before loading the strip on the unwinder at the entrance of the production line, and preferably before determining the production sequence of the strip.

-outputting data including the machining settings to an automation and/or control instrumentation system of the production line, and then performing the settings and generating the steel strip segment with the target mechanical, surface and/or geometrical properties.

By way of non-limiting example, the following issues illustrate the capability of DPPC:

for annealing, this setup takes into account the material chemistry, entry conditions and in-strip spatial inhomogeneities resulting from previous processing (hot rolling mill, pickling line, cold rolling mill, etc.) and the predicted processing speed and (de-) acceleration, defining the maximum and minimum temperatures determined by the required mechanical property specifications and the target temperature aimed at producing more uniform properties. The minimum and maximum temperature settings are also limited by factors such as the desired shape and surface properties prior to cladding processing and thermodynamics during strip transition.

For the processing (or strip) speed, the settings define target speeds, maximum and minimum limits, which are determined by the desired coating thickness and its control, the desired mechanical properties and its temperature-time control, the desired surface roughness, the temper rolling reduction and its control, etc.

For the cladding thickness, the settings define the pressure of the wiping gas medium, the distance between the gas jet outlets and the strip, the gas compressor head pressure, which are determined by the type and thickness of the cladding required, the required flattening reduction and straightening elongation, the intended machining speed and (de-) acceleration, etc.

For pre-galvanization surface conditioning, the settings define the combustion air-fuel ratio, the concentration of oxygen injected for oxidation, the dew point of the furnace atmosphere, etc., as determined by the chemistry of the material, the desired galvanizing bath chemistry, the predicted temperature change, the predicted process speed, etc.

For the strip shape before cladding, the settings define the temperature limits and maximum gradients of (rapid) heating and cooling, heating and cooling flux distribution over the strip width, line tension, out-of-flatness tolerance, etc., as determined by the strip dimensions (thickness to width ratio), feed roll dimensions, desired mechanical properties and temperature-time control thereof, predicted processing speeds, desired cladding thickness and control thereof, etc.

For temper rolling and/or stretch straightening, this arrangement limits the optimum reduction and/or elongation of the entire strip to maximize the original quality in view of the predicted annealing process, predicted process speed, predicted inline shape, material dimensions, desired mechanical, geometric and surface properties.

It should be noted that the incoming rolled strip may be a hot rolled strip or a cold rolled strip. Further annealing of the hot-rolled strip is usually not carried out, but may be necessary in certain cases. For example, if hot-rolled strip is to be hot-dip coated, the strip needs to be heated prior to immersion to avoid cooling the metal immersion bath below the operating temperature. This is commonly referred to as a heated cladding process. It is not necessary to heat the strip to influence the properties, for example in the recrystallization and/or austenitizing of the cold rolled strip. On the other hand, for hot rolled steel containing certain microstructures (e.g. martensite) or hot rolled steel that still has the potential for precipitation hardening due to dissolved elements, the annealing treatment may be beneficial to temper the martensite and improve the ductility of the steel or to promote precipitation and increase the strength of the steel.

In a preferred embodiment, the incoming steel strip is cold rolled steel. This is the most common form of annealing in the CPL according to the invention, and the manufacture of these steels must cope with increasingly strict tolerances and with increasing requirements on the number of allowed defects, mechanical properties, surface quality and dimensions and dimensional tolerances.

In a cyclic operation for real-time control applications, the process settings for a given production plan will be automatically calculated and sent to the automation and/or control instrumentation systems of the production line. If a goal violation is anticipated, an exception notification is generated and sent to the planning unit that generated the production plan.

The recalculation of the machining settings is preferably triggered automatically when the production schedule or input is changed or when the actual machining deviation exceeds the tolerance of the newly released machining settings due to unforeseen equipment failures, manual intervention in the machining speed, etc.

The invention is also embodied in a computerized process automation for a continuous production line, wherein the process automation is realized such that it executes the method according to the invention during operation.

The invention is also embodied in a continuous production line controlled by computerized machining automation according to the invention, and also in a continuous production line, wherein an FPS entering a steel strip section is determined, set and implemented during operation by CPL computerized machining automation according to the invention, to enable production of a steel strip section having targeted mechanical, surface and/or geometric properties.

The usefulness of the present invention is demonstrated by the following non-limiting examples and figures.

Figures 7 and 8 show measured values of mechanical properties of galvanized high strength steel grades produced from a continuous production line. Light grey indicates the results without DPPC. The black color represents the results for the same steel produced using a version of DPPC. The mechanical properties depicted are yield stress (Rp) and tensile strength (Rm). The performance without DPPC would show a wide distribution range with the measurement significance outside the tolerance range (dashed vertical line). 35% will be rejected on Rp (fig. 7) and 30% on Rm (fig. 8). DPPC predicts that the coiling temperature regime achieved in the run-out table cooling of the hot rolling mill (HSM) can be significantly improved by adjusting the annealing regime.

Controlled winding temperatures are used, wherein the head and tail of the hot rolled strip are wound at higher temperatures to compensate for the faster cooling rate at the outer wrap of the hot rolled coil. Fig. 5 shows the temperature of a large number of strips as a function of normalized length. These hot rolled and U-cooled strips were then pickled and cold rolled. In addition, DPPC proposes lower annealing temperatures for the head and tail of the cold rolled strips of these hot rolled coils compared to the intermediate section (or sections) of the strip, taking into account the spatial inhomogeneities in the strip caused by the previous processing steps. This is schematically illustrated in fig. 6. The horizontal line at about 705 c represents prior art practice of equal annealing temperatures throughout the length of the strip. The temperature profile is then applied to the cold rolled strip in the rapid heating furnace of the annealing step in said production line and the results of the final product are given in black in fig. 7 and 8. These values show a significantly reduced distribution width and better tolerance resistance. The rejection rate on Rp is reduced to 5%, and the rejection rate on Rm is reduced to 2%.