CN102354105B - The method and system of the multilamellar modeling of the substance characteristics in determining storage tank - Google Patents

Abstract

The invention discloses the method and system that the multilamellar of the substance characteristics in determining storage tank models.Using without in the batch processing control system of the storage tank of blender, storage tank can modeled and extract the characteristic of feed out to be precisely controlled the quality of process.Model need not measure injecting or extracting flow out or suppose most preferably to mix.But, improved model can assume that the feed being injected into storage tank still has the layering mixed with some because of the turbulent flow in Continuous convective, loading or other factors.This model can comprise the mapping of the characteristic of the storage tank layering of the injected material being described in this model.Storage tank being injected to each new loading of feed, the layer changed and feed is drawn out of can be updated by model layer according to the characteristic of described new loading.

Description

The method and system of the multilamellar modeling of the substance characteristics in determining storage tank

Cross-Reference to Related Applications

This application claims and benefit from disclosed in 21 days Mays in 2010, entitled " for determining substance characteristics in storage tank Multilamellar modeling method and system " U.S. Patent application No.61/347,208, at this by U.S. Patent application No.61/347, 208 are hereby incorporated by reference.

Technical field

The present invention relates generally to the invention of Process Control System, more particularly, to for determine without blender criticize at The realization of the multilamellar modeling technique of the flowing material characteristic in reason tank.

Background technology

Process Control System such as chemistry, oil or during other use Process Control System generally comprise communication Be couple at least one main frame or operator workstation and by simulation, numeral or combination analog/digital bus communication ground It is couple to one or more process controllers and input/output (I/O) equipment of one or more field apparatus.Field apparatus, It may be that such as, valve, valve positioner, switch and transmitter (such as, temperature, pressure and flow sensor).Those field apparatus Perform during the course such as to open or close the process control function of valve closing and measuring process control parameter.Process controller connects Receive the signal of process measurement representing that field apparatus is done, process this information to realize control routine, and generate control signal, This control signal is sent to field apparatus to control the operating of operation by bus or other order wires.In this way, mistake Range controller can perform and coordinate to use the control strategy of field apparatus via bus and/or other communication links.

The control information from field apparatus and controller can be made to can be used for by operator workstation (such as, based on process The system of device) one or more application (such as, software routines, program etc.) of performing, to enable the operator to implement about mistake The required function of journey, such as, check the current state (such as, passing through graphical user interface) of process, evaluation process, more correct one's mistakes The operation (such as, pass through visual object diagram) of journey, etc..A lot of Process Control Systems also include one or more application stations (example As, work station).In general, these application stations use and are communicatively coupled to controller, operation employee by LAN (LAN) Make the personal computer of other system in station and Process Control System, notebook or the like to realize.Each application station can To include graphical user interface, its display include process variable value and procedure correlation quality parameter value process control information, Procedure fault detection information and/or process state information.

Relate to the Process Control System of batch processing generally to process as a batch using the stage of various quantity or step Some general hyles or feed carry out producing product.The one or more steps of batch processing or stage can such as located The identical device of reason tank, reactor or other kinds of processing equipment performs.Feed is from other tanks of various such as storage tanks It is injected in reactor with different phase in batch processing in other reactors.From the scene being couple to storage tank and reactor Application that performed by operator workstation, one or more be may be used to guarantee operator's energy by the procedural information of equipment and controller Enough execution are about the function needed for batch processing.

For controlling the quality of batch processing product, accurately understand that each stage during the course there occurs that what is critically important 's.The characteristic understanding the feed being supplied to various process tank from storage tank in the different phase of process is to consider to determine final product One factor of quality.Such as, feed can have not according to comprising the many factors such as source, season, age, storage condition Same characteristic.Feedstock characteristic is not clearly understood from, controls the quality of final batch processing product by highly difficult.Such as, feed Any change of characteristic, even if within the acceptable range, can affect final batch processing product reactor operation and Mass parameter.

During some, when feed is drawn out of tank injecting reactor, when in order to for feed obtain uniform characteristic, When feed is passed to storage tank, feed is mixed.Under turbulent-flow conditions, mixed by convection current and turbulent diffusion.Diffusion can To be produced by distinct device: agitator tank, jet mixer and ultrasonic mixer.During feed mixes, it is also possible to logical Cross and measure the implantation concentration of storage tank, injection stream, extraction is flowed and storage tank level (or weight) realizes storage tank and extracts concentration out Continuous plus.Use the process of mixing feed it is also assumed that mix completely.

Batch operation is often used without mixing feed.The most periodically loaded in batch processing by truck or reactor Storage tank and do not use blender to mix feed.The transmission of additional feed can be along with analytical data, this analytical data Batch processing is allowed at least to illustrate to be added to the injection feed of previous feed.But, when storage tank removes, interpolation is more Feed.Inject feed and may comprise the characteristic the most different from old feed.Without in the storage tank of blender, when new feed quilt When adding old feed to, stratification or " layering " of a certain amount of feed will be produced.In addition to layering, due to abiogenous turbulent flow Also will produce the mixing of a certain amount of interlayer with other factors, but without blender in storage tank, it is impossible for being thoroughly mixed 's.Therefore, after feed leaves storage tank, very difficult prediction is injected into the precise characteristics extracting feed out of batch process reactor.

Summary of the invention

Using without in the batch processing control system of blender storage tank, the characteristic that can model storage tank extraction feed is come It is precisely controlled the quality of batch processing.Model need not measure injecting or extracting flow out or suppose most preferably to mix.And It is that improved model can assume that the feed being injected into storage tank still has because of the turbulent flow in Continuous convective, loading or other factors There is the layering mixed with some.This model can comprise storage tank layering (or layer) of the injected material being described in this model The mapping of characteristic.Storage tank being injected to each new loading of feed, change and feed are extracted to reaction by model layer The layer of device can be updated according to the new characteristic loaded (that is, this layer comprises extraction feed).

In certain embodiments, this model can comprise multiple stage.Such as, these stages can comprise by by old to The feedstock characteristic of the bed of material be applied to new to the bed of material to model tank, set up extraction layer according to storage tank horizontal survey, calculate storing Tank average characteristics, and calculate the outlet characteristic of extraction layer and recalculate average tank characteristic.Can independently or sequentially perform These stages.

In other embodiments, determine that the feed material characteristic in the storage tank without blender of process control plant is permissible It is applied to, newly to the bed of material, export, based on tank, the storage tank horizontal survey being associated to the feedstock characteristic value of the bed of material including by previous Set up extraction layer, and calculate the meansigma methods of the feedstock characteristic of total feeding coal in described storage tank.Can also calculate described Extract the hybrid cytokine of layer out, and meansigma methods based on described feedstock characteristic and described hybrid cytokine can calculate described extraction The extraction feedstock characteristic value of layer.Feed in described extraction layer mixes with other feed section in described storage tank.

In other embodiments, the extraction feed material in the storage tank without blender that batch processing is calculated or determined The method of characteristic includes the meansigma methods calculating the feedstock characteristic value of the total feeding coal in storage tank.Described total feeding coal can comprise New to the feed in the bed of material.The method is the described new position giving the bed of material and the position of extraction layer in further comprising determining that tank.Institute The position stating the new position to the bed of material and extraction layer is respectively corresponding to the entrance and exit of storage tank.The method can also include base Extraction feedstock characteristic value is calculated in hybrid cytokine.Described hybrid cytokine can comprise the value of the feedstock characteristic value that experiment draws and subtract The feedstock characteristic value going nothing mixing deducts the ratio of the described feedstock characteristic value without mixing with the feedstock characteristic value being thoroughly mixed The slope of the regression line of the optimal coupling of discrete figure.

In other embodiments, in a kind of computer installation may determine that the storage tank without blender of process control plant Feed material characteristic.Described device comprises: that be stored thereon, comprise the computer of the computer-implemented application of some routines Readable memory.Such as, the first routine can detect storage tank new feed transmission and the second routine can with corresponding to The feedstock characteristic value more new data structure of this new feed transmission.3rd routine can calculate the feed of the total feeding coal in storage tank The meansigma methods of characteristic value, wherein total feeding coal comprises new to the new feed transmission in the bed of material.In 4th routine may determine that tank The described new position giving the bed of material and the position of extraction layer, and the 5th routine can calculate extraction based on the hybrid cytokine extracting layer out and give Material characteristic value, described hybrid cytokine comprises the slope of the regression line of optimal coupling described here.

Accompanying drawing explanation

Figure 1A show can be used for the process that realizes, there is controller and the showing of the process control network of field apparatus It is intended to；

Figure 1B shows Process Control System exemplary including exemplary operation management system and multilamellar MBM Block diagram；

Fig. 2 shows the data structure for the exemplary batch comprising process variable and quality variable；

Fig. 3 shows the data structure of the multiple exemplary batch for comprising process variable and respective quality variable；

Fig. 4 shows for multilamellar modeling to determine the storage tank of the substance characteristics in storage tank；

Fig. 5 shows the data structure of tissue storage tank material characteristic data；

Fig. 6 A shows the complex function block for the substance characteristics being used together to determine in storage tank with multilamellar modeling；

Fig. 6 B shows the functional device view of the complex function block shown in Fig. 6 A；

Fig. 7 A shows the data for calculating hybrid cytokine；

Fig. 7 B shows the discrete figure of data in Fig. 7 A；

Fig. 8 shows input in time and extracts the chart of characteristic value out；

Fig. 9 shows the flow process of the method and system of the multilamellar modeling of the substance characteristics described in determining storage tank Figure；

Figure 10 shows that can be used for realizing multilamellar described here, substance characteristics in determining storage tank builds The illustrative methods of mould and the block diagram of the example processor system of system.

Detailed description of the invention

Exemplary method described here and device may be used for Process Control System, special with the flowing material during providing Property so that when process occurs or carries out, operator can makeover process fault.Such as, method and apparatus described here Can with U.S. Patent Application No. No.12/538,995 disclosed in 11 days Augusts in 2009 described in operation management system (OMS) it is used together, is hereby incorporated entire contents.Process can be performed in response to the flowing material characteristic that storage tank is extracted out Revise.Further, it is determined by the flowing material characteristic of the extraction of the storage tank without blender and adjusts the downstream of batch processing Reactor and other processes, illustrative methods described here and device may be used for revising product quality.Mistake described here Process control system can comprise any type in batch processing system, continuous processing system, automatic system and/or manufacture system.

Figure 1A shows example process control system 10, including being connected to data history records 12 and one or more Main work station or the process controller 11 of computer 13 (it can be any type of personal computer, work station etc.), each Main work station or computer 13 have display screen 14.Controller 11 blocks 26 and 28 also by input/output (I/O) and is connected to now Field device 15-22, and the one or more batches operation using field apparatus 15-22 to realize batch processing can be run.Number Can be the data collection module of any desirable type according to historical record 12, it has the memorizer of any desirable type and appoints Needed for meaning or known for storing the software of data, hardware or firmware.Data history records 12 can with in work station 13 Separation (as shown in Figure 1A), or work station 13 in the part of.Controller 11, it can be such as The DeltaV that Emerson Process Management is soldController, is connected or any other institutes by such as Ethernet The communication network 23 needed is communicatively connected to master computer 13 and data history records 12.Controller 11 uses and such as marks Quasi-4-20 milliampere equipment and/or such as FOUNDATIONFieldbus agreement, HARTAgreement, WirelessHART^TMAgreement etc. The arbitrarily required hardware and software that is associated of any smart communication protocol and be communicatively connected to field apparatus 15-22.

Field apparatus 15-22 can be any type of equipment, such as sensor, valve, transmitter, localizer, etc., and I/ O card 26 can be to meet the most required communication or any type of I/O equipment of controller protocol with 28.In the reality shown in Figure 1A Executing in example, field apparatus 15-18 is the standard 4-20 milli that the simulation by artificial line or combination communicates with I/O card 26 with digital line Install standby or HART device, and field apparatus 19-22 is such as FOUNDATIONThe intelligence of Fieldbus field apparatus sets Standby, it uses Fieldbus communication protocol to be communicated with I/O card 28 by number bus.Certainly, field apparatus 15-22 can meet Arbitrarily other required one or more standards or agreements, including the arbitrary standards developed in the future or agreement.

Controller 11 includes the process realizing or monitoring one or more process control routine (storage is in memory 32) Device 30, it can include controlling loop, and communicate with equipment 15-22, master computer 13 and data history records 12, thus Process is controlled in the way of the most required.It should be noted that, if it is desired, arbitrary control routine described herein or module are permissible Have by the different parts that controller realized or performed.Similarly, described herein will be in Process Control System 10 The control routine or the module that realize can be to use arbitrary form, including software, firmware, hardware etc..Can be with arbitrarily required Software format realizes control routine, such as use OOP, use ladder logic, sequential functional diagram, FBD or Person uses arbitrarily other software-programming languages or design example.Similarly, control routine can be arrived by hard coded, and such as one Individual or multiple EPROM, EEPROM, special IC (ASIC) or arbitrarily in other hardware or firmware components.Therefore, may be used By Configuration Control Unit 11 for realizing control strategy or control routine in the way of the most required.

In certain embodiments, controller 11 use commonly called functional device to implement control strategy, each of which merit Energy block is object or other part (such as subroutine) of overall control routine, and with other functional devices (by being referred to as The communication of link) synthetic operation to be to realize Process control loop in Process Control System 10.Functional device generally implements input work The one that can, control in function or output function to perform some physical functions in Process Control System 10, this input function The function being such as associated with transmitter, sensor or other process parameter measurement device, this control function be such as with reality Executing the function that PID, fuzzy logic etc. control to be associated, this output function is the merit of the operation of some equipment controlling such as valve Energy.Certainly, there is also mixing and other kinds of functional device.Functional device can be stored in controller 11 and by it Performing, this is typically to be used in these functional devices, or is associated with standard 4-20 milliampere equipment and such as HART device In the case of some type of smart devices, or field apparatus itself can be stored in and by it in these functional devices Realizing, this can be in the case of relevant Fieldbus equipment.

As shown in the block 40 of the decomposition of Figure 1A, controller 11 can include some monocycle controls as shown in routine 42 and 44 Routine processed, and if it is desired, it is possible to realize one or more Dynamic matrix control loop, such as control shown in loop 46 many Individual/input-multiple/output control routine.Each this loop is commonly called control module.Monocycle control routine 42 and 44 quilt Be shown as using simulation input (AI) that single input/single output fuzzy logic control block and being respectively connected to is suitable for Single input/single output the PID control block of simulation output (AO) functional device performs monocycle control, and these control block and can close It is coupled in measurement equipment or the Process Control System 10 of the process control equipment of such as valve, such as temperature and pressure transmitter Other equipment any.Dynamic matrix control loop 46 is shown as including being communicably connected to the defeated of one or more AI functional device Enter and be communicably connected to the output of one or more AO functional device, but the input of advanced control block 48 and output can be by It is connected to arbitrarily other required functional devices or control element to receive other kinds of input and to provide other kinds of control Output.Advanced control block 48 can be module PREDICTIVE CONTROL (MPC) block, backbone network models or controls block, Multivariable Fuzzy is patrolled Collect any type controlled in block, real-time optimization block etc., or can be the control block etc. adjusted with being suitable for.It is appreciated that Functional device shown in Figure 1A, including advanced control block 48, it is possible to performed by controller 11, or alternatively, it is possible to it is positioned at such as In other processing equipments of in work station 13 one in one or even field apparatus 19-22 and be executed by.

Additionally, as shown in figure ia, one or more process analysis routines 50 can each by Process Control System 10 The equipment of kind stores and performs.Although process analysis routine 50 is illustrated as stored in one or more computer-readable memory 52 In to be performed on the processor 54 of work station 13, routine 50 can alternatively be stored in and executed in other equipment.Often Individual process analysis routine 50 is communicatively coupled to one or more control routines of such as control routine 42,44,46, and/or It is coupled to data history records 12 and measures to receive one or more measured process variable.Each process analysis routine 50 Can be used for developing statistic processes model and analyzing ongoing or online batch processing according to this model.Analysis routines 50 is also Can show about online or carry out the information of middle batch, such as Process Control System 10 institute to the user of such as batch operation person As Shi Xianing.In certain embodiments, process analysis routine 50 can include determining described here from without mixing The routine of the substance characteristics of the feed output of the storage tank of device.

Figure 1B shows the block diagram of another example of process control environment 100, and this process control environment 100 includes running pipe Reason system (OMS) 102, its be otherwise known as process monitoring and quality prediction system (PMS).OMS 102 is positioned at and includes process control In the factory 104 of system 106.Exemplary factory 104 can be any type of manufacturing facility, process equipment, automation installation, And/or any other kinds of process control structure or system.In some instances, factory 104 can include being positioned at not coordination The multiple facilities put, although and the factory 104 of Figure 1B be shown as including that Process Control System 106, factory 104 can also be wrapped Include additional Process Control System.

Process Control System 106, it is communicatively coupled to controller 108 by data/address bus 110, this process control system System 106 can include any number of field apparatus (such as, input and/or outut device) for realizing process function, should Process function such as perform during physical function or carry out the measurement of process variable.Field apparatus can include any type Process control modules, this process control modules be able to receive that input, generate output and/or control process.Such as, scene sets For including that the input equipment of such as valve, pump, fan, heater, cooler and/or blender is with control process.Additional Ground, field apparatus can include the defeated of such as thermometer, piezometer, densitometer, liquidometer, effusion meter and/or gas sensor Go out equipment to the process variable measuring in process or in partial routine.Input equipment can receive instruction to hold from controller 108 The one or more orders specified of row also cause process change.Additionally, outut device measure process data, environmental data and/or Measured data are also sent to controller 108 as process control information by input equipment data.This process control information Value (such as, the measured process variable corresponding to the variable from the output measured by each field apparatus can be included And/or measured quality variable).

In the example shown in Figure 1B, controller 108 can be by data/address bus 110 and Process Control System 106 Field apparatus communicates, and it can be coupled to the intermediate communication assembly in Process Control System 106.These communications components can wrap Include field terminal box so that the field apparatus in command area is communicatively coupled to data/address bus 110.Additionally, communications component is permissible The communication path of field apparatus and/or field terminal box is organized including scheduling cabinet.Additionally, communications component can include I/O Block to receive data from field apparatus and convert data to the telecommunication media that can be received by example controller 108.This Data from controller 108 can be converted to the data form that can be processed by corresponding field apparatus by a little I/O cards.One In individual example, (such as Profibus assists can to use Fieldbus agreement or other kinds of wiredly and/or wirelessly communication protocol View, HART protocol etc.) realize data/address bus 110.

The controller 108 of Figure 1B manages one or more control routine, thus the scene in management process control system 106 Equipment.Control routine can include process monitoring application program, alarming and managing application program, process trend and/or historical usage Program, batch processing and/or activity management application program, statistics application program, stream video application, Dynamic matrix control application program Deng.Additionally, process control information can be transmitted to OMS 102 by controller 108.Control routine can be implemented to guarantee process Control system 106 manufactures the required product in certain mass threshold value of specified amount.Such as, Process Control System 106 can be by It is configured at the end of batch manufacture the batch system of product.In other example, Process Control System 106 can include not The continuous process manufacturing product manufactures system disconnectedly.

Come from the process control information of controller 108 can include corresponding to measured, derive from process control system The process of field apparatus and/or the value of quality variable in system 106.In other examples, OMS 102 can be by process control information In value resolve to corresponding variable.Measured process variable can be associated with process control information, this process control information Derive from measurement partial routine and/or the field apparatus of field apparatus characteristic.Measured quality variable can be associated with process Control information, this process control information is about measuring the characteristic with the process being associated at least partially completing product.

Such as, process plant can be changed in the tank forming certain density chemicals in a fluid or reactor Learn reaction.In such examples, in fluid, the concentration of chemicals can be a kind of quality variable.The temperature of fluid and fluid stream The speed entering tank can be process variable.Being modeled by process control and/or monitor, OMS 102 may determine that fluid in tank On the basis of concentration is built upon the temperature of fluid in tank and fluid (such as, the feed) flow velocity of inflow reactor.Therefore, Not only concentration is quality variable, and rate of flow of fluid and fluid temperature (F.T.) all can act on or affect the quality of concentration.In other words, surveyed The process variable of amount acts on or affects the quality of measured quality variable.OMS 102 can use statistical disposition to come really Determine the impact on quality variable of each process variable and/or the degree of effect.

Additionally, OMS 102 can model and/or determine that the measured process being associated with Process Control System 106 becomes Relation between amount and/or quality variable.Measured these relations between process and/or quality variable can form one Or the multiple quality variable being computed.The quality variable being computed can be one or more measured process variable, be surveyed The quality variable of amount and/or the combination of the multivariable and/or linear algebra of other quality variables being computed.Additionally, OMS 102 can be determined always by the combination of measured process variable, measured quality variable and/or the quality variable being computed Weight variable.Oeverall quality variable can determine corresponding to the quality of whole process and/or can being produced corresponding to process The quality of the prediction of raw product.

The OMS 102 of Figure 1B includes analysis processor 114, and it uses descriptive modelling, prediction modeling and/or optimizes next life Become about Process Control System 106 state and/or the feedback of quality.Analysis processor 114 can be communicatively coupled to modeling One or more MBMs 115 of the various characteristics of the process performed by factory 104.Analysis processor 114 can detect, really Determine and/or diagnose process operation fault and predict that Arbitrary Fault is to being associated with the matter of product produced by Process Control System 106 The quality variable of amount and/or the impact of oeverall quality variable.Additionally, analysis processor 114 can pass through statistically and/or logic Quality and/or process variable are combined to the oeverall quality variable of the oeverall quality of the process that is associated with to monitor the matter of process by ground Amount.Analysis processor 114 then can by the value calculated for oeverall quality variable and/or be associated with the value of other quality variables with Each threshold value compares.These threshold values can set up the predetermined matter of the oeverall quality variable at different time during the course On the basis of amount limit.Such as, if the oeverall quality variable being associated with process exceedes threshold value a period of time, then predicted The final mass of produced product may not reach the quality metric being associated with final products.The one of MBM 115 Individual example comprises feedstock characteristic MBM 115, and its modeling is described here without the substance characteristics in the storage tank of blender.

If oeverall quality variable and/or arbitrarily other quality variables deviate respective threshold value, then analysis processor 114 is permissible In process observation figure and/or process variation graph generate indicating fault, its display explained and/or do not explained with totally The deviation (or change) that quality variable is associated, and/or the variable producing this procedure fault can be shown.Exemplary analysis processes Device 114 by offer enable an operator to generate can show measured process variable, measured quality variable and/ Or current and/or procedure quality figure (such as, constitutional diagram, mini figure, the process variations of past value of the quality variable etc. being computed Figure, variable trend graph, image etc.) function carry out management analysis, to determine the cause of one or more procedure fault.Additionally, point Analysis processor 114 generates these figures when process operation, and when OMS 102 receives additional process control information, point Analysis processor 114 is continuously updated and/or recalculates the multivariate statistics being associated with each figure.

Analysis processor 114 can by calculate process variable and/or quality variable relative to oeverall quality variable or Cause the distribution of quality variable of fault to generate scattergram.The distribution of process and/or quality variable can be shown as being solved The deviation of each variable that is that release and/or that do not explained, and be associated with the deviation of oeverall quality and/or be associated with the matter of fault The distribution of quantitative change amount.

Additionally, example analytic processor 114 can be arbitrarily chosen process and/or quality variable generates variable and becomes Gesture figure, this variable is likely to be of the deviation more than predetermined threshold.Variable trend graph can show a period of time of the process of being associated with On the value of variable, this value be related to before during similar temporal variate-value.Become by generating scattergram and/or variable Gesture figure, analysis processor 114 can also identify feasible revising process with the fault detected by solution.Had by offer Having the overlap of the historical record figure of the deviation with currency (such as standard deviation) being associated, variable trend graph can help Operator determines the cause of procedure fault.

Analysis processor 114 can also generate quality prediction graph to determine, if realized, one or more corrections are right Effect in process oeverall quality.If oeverall quality is maintained or is improved in the threshold value specified, then by one or more corrections Analysis processor 114 may indicate that OMS 102 realizes one or more correction.Alternatively, analysis processor 114 can be to control Device 108 sends instruction with the one or more corrections realizing process.

Further, once it is determined that be associated with the fault of oeverall quality variable and/or any other quality variables, exemplary Analysis processor 114 can generate mini figure.Mini figure can include the meansigma methods about each variable and/or standard deviation At the appointed time (such as, be associated with the time of procedure fault) on process and/or the value of quality variable.Additionally, mini Figure can include the spark line pointing out to be associated with the preferred value of each process and/or quality variable.From mini figure, exemplary point Analysis processor 114 is so that operator can determine and/or selects one or more correction to process operate and/or determine Arbitrary revise whether by development so that oeverall quality variable is predicted in specified limit.

Exemplary OMS 102 manages access and the control of process control data, this process by online data processor 116 Control data and include process variations figure, scattergram, variable trend graph, quality prediction graph and/or mini figure.Additionally, count online There is provided access to check that process control data, change and/or change procedure control number according to processor 116 to Process control operators According to and/or generate the instruction of field apparatus in the Process Control System 106.

The factory 104 of Figure 1B includes the road being communicatively coupled to online data processor 116 by LAN 124 (LAN) By 120 and local work station 122.Additionally, other the work station any in factory 104 can (not shown by exemplary route 120 Go out) it is communicatively coupled to LAN 124 and/or online data processor 116.Route 120 can be wirelessly and/or by wired company Connect and be communicatively coupled to other work stations.Route 120 can include any type of wireless and/or wired route, as connection Access hub to LAN 124 and/or online data processor 116.

The most required telecommunication media and protocol realization LAN 124 can be used.Such as, LAN 124 can be based on rigid line Or wireless ethernet communication plan.It is also possible, however, to use arbitrarily other be suitable for telecommunication media and agreements.Although additionally, showing Go out single LAN, but the communication hardware being suitable in more than one LAN and work station 122 can be used to come at work station 122 With enough communication path is provided between respective similar work station (not shown).

LAN 124 also can be communicatively coupled to fire wall 128.This fire wall 128 comes according to one or more rules Determine whether the communication from remote work station 130 and/or 132 is allowed to access factory 104.Exemplary remote work station 130 With 132 access that can provide resource in factory 104 to the operator in being not in factory 104.Remote work station 130 He 132 are communicatively coupled to fire wall 128 by wide area network (WAN) 134.

Exemplary workstation 122,130 and/or 132 can be configured to observation, change and/or makeover process and control system One or more process in system 106.Such as work station 122,130 and/or 132 can include user interface 136, its layout And/or the process control information that display is generated by OMS 102.Such as, user interface 136 can receive from OMS 102 and be generated Figure and/or chart, or alternatively, receive the data for generating process control graphic and/or chart.Once at each Receiving figure and/or chart data in work station 122,130 and/or 132, user interface 136 can generate and relatively be easier to as behaviour Figure that work person is understood and/or the display of chart 138.Example in Figure 1B shows the work station with user interface 136 132.But, work station 122 and/or 130 can include user interface 136.

User interface 136 can control operator with reminder process and note in Process Control System 106 and/or factory 104 The arbitrarily arbitrary process in other Process Control System controls the generation of fault.Additionally, user interface 136 can guide control Operator determines the source of procedure fault by analysis process and predicts the procedure fault shadow to produced product quality Ring.User interface 136 can provide process control statistical information when process is carried out to operator, so that operator can It is adjusted process revising arbitrary fault.By the fault in makeover process, operator can keep produced product The quality of product.

By exemplary OMS 102, exemplary user interfaces 136 can show detection, analyze, revise operation and Quality prediction information.Such as, user interface 136 can show process overview chart, process variations figure, mini figure, scattergram, can Become trendgram and/or quality prediction graph (such as, figure 138).Once observing these figures 138, operator can select to add Figure 138 observe multivariate and/or statistic processes information to determine the cause of procedure fault.Additionally, user interface 136 Can show for makeover process fault, feasible operation.User interface 136 can after allow operator select one or Multiple correction operates.Once selected revising, correction can be sent to OMS 102 by user interface 136, then, OMS 102 again to Controller 108 sends instruction, to make applicable correction in Process Control System 106.

The exemplary workstation 122,130 and/or 132 of Figure 1A can include calculating equipment arbitrarily, and it includes individual's electricity Brain, notebook computer, server, controller, personal digital assistant (PDA), microcomputer, etc..Can use arbitrary Be suitable for computer system or processing system (such as, the processor system P10 of Figure 10) realize work station 122,130 and/or 132.It is for instance possible to use the work station etc. of the PC of single processor, single or multiple processor realizes work station 122,130 and/or 132.

Example process controls environment 100 and is provided to a type of the system that illustrates, describe more particularly below shows The method and device of example can be used the most wherein.If however, if it is desired to, exemplary side described herein Method and device can be advantageously used in complexity and control environment 100 higher or lower than the example process shown in Figure 1A And/or in the other system of Process Control System 106, and/or be used in cohesive process control activity, enterprise management activity, In the system that communication activity etc. use.

Fig. 2 depicts the data structure 200 for exemplary batch (such as, batch #1), and it includes measured variable 202 and the quality variable 204 that is computed.Example data structure 200 can also comprise oeverall quality variable (not shown), and logical Cross measurement or observation, this oeverall quality variable can be obtained at the end of batch.Batch processing is a type of product manufacturing, In this product manufacturing type, create larger amount of product and/or product concurrently in the one or more positions controlled by routine Part.This routine can comprise one or more process stage, and each stage comprises one or more operation, and each behaviour Make to comprise one or more phase.When exemplary method described here and device refer to batch processing, it is possible to achieve any class The process of type.

Exemplary measured variable 202 comprises measured process and/or quality variable.Such as, variable P1 corresponding to The flow velocity (such as, process variable) of fluid, and variable P2 is corresponding to the concentration (such as, quality variable) of fluid.In conjunction with batch processing Batch #1 show measured by variable 202.Batch processing occurs along (such as, time) in the time cycle that t-axle shows.Additional Ground, the batch processing in Fig. 2 comprises 8 measured variablees.But, in other example, batch processing can comprise less or more Many measured variablees.

It is relevant to the only number of times during batch processing that Fig. 2 also show some measured variablees 202.Such as, Variable P1 is relevant to the midpoint of the starting point from this batch to this batch.Therefore, if variable P1 is associated with the flow velocity of fluid, then stream Only starting point in this batch is flowed by body to during the batch processing at midpoint.After midpoint, this batch is by the most recycling fluid stream Amount, therefore, variable P1 does not associates with the batch processing crossing midpoint.On the contrary, variable P4 associates with whole batch processing.

When exemplary quality variable 204 association being computed maybe may be associated with batch processing specific with whole batch processing Phase or stage.The quality variable 204 being computed can be changeable between measured variable 202 and/or other quality variables 204 The result of amount, statistics and/or algebraically relation.Such as, quality variable Q1 204 can be corresponding to produced in batch processing The composition quality of product.Composition quality Q1 is a quality variable, because it is not directly to measure in Process Control System 106 's.On the contrary, according to the multivariable combination of measured variable 202 P1, P3, P4 and P7, can model and/or determine composition Quality Q 1.Therefore, any one if composition quality Q1 exceedes predetermined threshold, then in measured variable P1, P3, P4 and P7 And/or the combination of measured variable P1, P3, P4 and P7 all facilitates factor by can become deviation.

Fig. 3 shows data structure 300, and what it can be used for comprising process variable 302 and respective quality variable 304 one is Arrange exemplary batch.These batches (such as, batch 1-7) show that batch processing contains the stage being performed with serial order (such as, stage 1-4).Such as, the stage 1 can be corresponding to chemical substance in batch or the combination of other feeds and mixing, and rank Section 2 is corresponding to the process of those mixed chemical substances in batch.These stages can further be subdivided into operation, time Phase and/or level.Additionally, the quality variable 306 being computed in each batch corresponding to measured variable 302.

Fig. 4 shows exemplary storage tank 400, and this storage tank can be used for certain single order in batch processing as described in Figure 3 Section, (such as, stage 1).As detailed above, determine that feed is extracted characteristic out and be can aid in fault for the feed 402 in storage tank 400 Detection and the quality of batch processing product.In certain embodiments, the feedstock characteristic modeling of operation management system (OMS) can be passed through Module 115 (Figure 1B) performs for determining that feed extracts the function of characteristic, instruction, method out.

When this process is online, it is used for supplementing process monitoring, fault detect, prediction of quality and process control, only arranges When having lifted some potential application on site, it is possible to implement described here, for determining the calculating of feedstock characteristic.Can also pass through For supplementing or verify line modeling off-line procedure, the exploitation detection model in principal component analysis (PCA), partial least square method Analyses etc. are implemented described here, for determining the calculating of feedstock characteristic.

Without the storage tank 400 of blender comprise feed 402 be passed to the entrance 404 of storage tank 400 and feed 402 by from Storage tank 400 transports hence into the outlet 406 of (such as, reactor 407 or other assemblies) in another assembly of factory.Such as Fig. 4 Shown in, storage tank 400 can comprise the different configurations of entrance 404 and outlet 406.Such as, storage tank 400 can comprise top and enters Mouthful 404a, centre entrance 404b or bottom inlet 404c and top exit 406a, centre exit 404b or outlet at bottom 404c's Any combination.Certainly, entrance 404 and outlet 406 described in Fig. 4 are for illustrating common region, entrance 404 and outlet 406 Can be placed on tank 400.Although any point that entrance 404 and outlet 406 can be placed in storage tank 400, with complete Becoming feed 402 to the transmission back and forth of storage tank, method described herein can comprise the default location of outlet at bottom 406c.

Feed 402 may determine that stream type to the situation that storage tank 400 is transmitted back and forth, and the type may result in storage tank The degree of the mixing of the various transmission of the feed 402 in 400.Such as, comprise bottom inlet 404c when storage tank 400 and top goes out During mouth 406a, it can be assumed that the mixing of high level and stream type can be described as " being thoroughly mixed ".But, when storage tank 400 When comprising top entry 404a and outlet at bottom 406c, it can be assumed that the mixing of low degree and stream type can be described as " plug Shape stream ".Slug flow can describe export environment, and in this environment, feed is with minimum admixture and every time by from storage tank Extraction in the layer of the bottom of 400.When entrance 404 and outlet 406 are placed on the same point in storage tank 400 (along with one During a little feed transmission, some deviations by caused by the quantity of the feed 402 in storage tank 400), stream type can be described as " short-circuit ".

When stream type is not thoroughly mixed, once new feed is passed to tank 400, it will produce certain of feed 402 The stratification of degree or layering.The feed to storage tank 400 in discrete time interval transmits to create to be had at each layer Between the level of various transmission of some degree of feed mixing or " layer ".With reference to Fig. 4, feed transmission may result according to from the oldest Transmission 408 is to the layer that transmission 416 arranges recently.Therefore, when storage tank 400, currently there are five discrete transmission of feed 402 Time, storage tank 400 contains five layers of feed 402, and it is as pointed by the layering 408,410,412,414 and 416 in Fig. 4. The chemistry, physics and other reactions that are caused by the various concentration between different transmission and other differences can cause feed to be layered The slight mixing of 408-416.Such as, because of in storage tank moves, entrance and exit moves, flow etc. causes storage tank 400 The movement of feed can cause the generation of mixing.

The transmission each time of the feed 402 of storage tank 400 can comprise multiple substance characteristics.Fig. 5 shows that data are tied The diagram of structure 500, this data structure 500 can be communicatively coupled to feedstock characteristic MBM 115, and be used for storage The characteristic 502 of the feed 402 currently having in storage tank 400.Characteristic 502 can include physics, chemistry and other feedstock characteristic Value 504, it includes pH balance, degree of reaction, toxicity, concentration, density, molecular weight and viscosity, only lists.Feed 402 The characteristic of transmission can be contained in from the data that feed suppliers receives every time, and it also can be come by the analysis of receptor Determine or by other suitable methods, operation management system (OMS) 102 can be passed to.Quantitative characteristics 506 comprises expression institute The value of the volume of feed, weight or the quantity of transmission.(such as, quantitative characteristics 506 can be used for the physical characteristic with storage tank Storage tank capacity, inside dimension, storage volume restriction etc.) combine, thus determine other measured value, such as, storing In tank each to the Position Approximate of the bed of material, the Position Approximate of the Mixed Zone between each layer, etc..Time response 508 indicates When feed is passed to storage tank.Time response also can be used to determine other characteristics of feed.Such as, some feedstock type Characteristic 502 may time to time change in a known way.When the special layers of feed is aged, seasoning layer can be changed Characteristic value 504 thus reflect any of change.Similarly, due to the different chemically and physically spy of the feed in each layer Property, mixing may increase over time.Similarly, hybrid cytokine described herein can change to reflect due to old Change, mixing and other factors caused by change.

In certain embodiments, data structure 500 may be implemented as stack data structure, the feed in its reflection storage tank Layer position and the position relative to entrance 404 with outlet 406.Such as, when outlet acquiescence bottom position 406c (Fig. 4) and Entrance is when tip position 404a, and now data structure may be implemented as " first in first out " (FIFO) stack, the oldest feed Transmission is extracted out at first in tank, and the transmission of up-to-date feed is finally extracted out in tank.It is similar to, when outlet position bottom acquiescence Putting 406c and entrance also when bottom position 404c, now data structure may be implemented as " first-in last-out " (FILO) stack, Which depict following configuration, the transmission of the oldest feed is finally extracted out in tank, and the transmission of up-to-date feed is taken out at first in tank Go out.When outlet is when giving tacit consent to bottom position 406c and entrance at centre position 404b, it is possible to use FIFO and FILO stack method Mixing.Stack 500 may be implemented as matrix, chained list or other kinds of known data structure.

In operation, the stack 500 physical state with the feed 402 in reflection tank 400 can be updated.Such as, when new feed 402 when being passed to tank 400, uses one or more programmed instruction that series of characteristics 502 is pressed into stack 500.If one Layer is fully removed in tank, then other programmed instruction " can eject " in stack and remove characteristic.Other instructions can be real Existing pointer 510 is to activate or current just by extraction in tank 400 (that is, extracting layer out) to indicate which layer.Such as, at slug flow In the tank 400 of configuration, extracting layer out and be always layer 408, until layer 408 is removed, then layer 510 becomes activation.At this example In, pointer 510 will indicate the characteristic of " layer 1 " always, until layer 1 is removed.Once removing, the characteristic of layer 1 will eject in stack, And the characteristic of previous layer 2 is it will be assumed that position is in stack top (that is, layer 1).When stack 500 is implemented as FIFO stack, and pointer 510 will refer to Series of characteristics to stack 500 top (such as " top " or " peek " instruct).In operation, when a layer is by tank 400 When removing, pointer 510 will be switched to the next layer that will be drawn out of.When the position of the entrance 404 in tank 400 and outlet 406 is wanted The realization of FIFO stack 500, pointer 510 is asked to will correspond to old layer, as by indicated by time response 508, until quantity 506 remove in tank 400, and then pointer 510 will be switched to next old layer, then repeat.

Feedstock characteristic MBM 115 can be configured to complex function block be used together with OMS 102 determine by from The characteristic of the feed 402 of reactor 407 or other process plant entities it is retracted to without the storage tank 400 of blender.Fig. 6 A illustrates For determining the example of the complex function block 600 of feedstock characteristic.Fig. 6 B shows the functional device view of complex function block 600 The example of 650.Functional device view 650 will be present in user interface application, graphically present Process Control System 100 Any aspect is also revised or other any aspects of control function block 600 or Process Control System 100.In certain embodiments, User interface is DeltaVControlStudioApplication.Returning to Fig. 6 A, complex function block 600 can comprise one or more merit Can block, such as computing block 604 and other functional devices.Complex function block 600 can limit tank property calculation block 604 and comprise multiple The parameter limiting feedstock characteristic 502 and other parameters of the deviation comprised in storage tank 400 design.Therefore complex function block 600 Stack data structure 500 can be communicatively coupled to.Further, complex function block can be configured, and comes automatically or passes through User interface updates stack 500.

In certain embodiments, parameter comprises description storage tank 400 currently to the bed of material or the outside of the total number of level 608 Parameter (cumulative volume of feed in such as tank), to storage tank 400 feed entrance 610 (such as top, centre, bottom) point, Hybrid cytokine 612 with each layer mixability described in storage tank 400.Hybrid cytokine 612 can include the value from 0 to 1, its In 0 mean without mixing and 1 means to be thoroughly mixed.Generally, it is configured at the feed-stock outlet of bottom with in centre when tank 400 Or during the entrance of top entry position 404a, the hybrid cytokine of slug flow tank will be greater than 0.Additionally, when tank 400 was configured to the end of at The feed-stock outlet in portion and when the entrance of bottom inlet position 404c, the hybrid cytokine of slug flow tank will be close to 0.Hybrid cytokine 612 can be used to whole tank 400 or for each layer in tank or level or describe the one or more of feed for combining of layer The mixing situation of characteristic.

Data parameters can also include multiple external input parameter, and it is for describing the feed of the specified level in tank 400 The input characteristics 614 of 402, and externally ported characteristic 616, it is for being described in (that is, taking out to the bed of material of the level of outlet 406 Go out layer) characteristic.Feed input characteristics 614 and outlet characteristic 616 can include indicating feed chemistry, physics and other characteristics Multiple values (such as, PH balance, degree of reaction, toxicity, concentration, density, molecular weight etc.).Input characteristics 614 can comprise when feed quilt The value measured when being input to storage tank 400 or suppose, and export characteristic 616 and can comprise for just being extracted out from storage tank 400 The value calculated by tank property calculation block 604 to the bed of material.As it has been described above, with reference to Fig. 5, complex function block 600 can be visited from stack 500 Ask input characteristics 614.Although Fig. 6 A shows seven inputs and outlet characteristic 616, but may have less or greater number Characteristic.Other blocks can comprise add block 618 with instruction be added to storage tank feed amount (such as weight, volume or other The measurement of the amount of the feed of the certain layer being added in storage tank) and date/time block 620 comprise description specific to bed of material quilt Add the value of the date and time of storage tank 400 to.

Functional device can comprise output parameter 622, putting down of each tracked characteristic of its total feed in index pot Average, and comprise stack pointer output 624, it is used for which indicates just extracted out and pointer 510 in tank 400 to the bed of material is current Current location.Tank property calculation block 604 can comprise one or more instruction being executed by processor or formula, for base Outlet characteristic 616 is determined in input characteristics 614 and other above-mentioned data.In certain embodiments, quilt is newly transmitted when feed 402 When being input to storage tank, stack 500 is updated and new feedstock characteristic " is pressed into " in stack 500.Such as, update stack to comprise and add Add a new range characteristic to add a number of feed 506 and new layer 512 with explanation to stack 500, increase horizontal block 608, arrange Interpolation on external input parameter 618, with the amount of the feed 506 that reaction is included in new transmission, and is set in outside defeated Enter the date/time in parameter 620 to react this newly added time 508.Tank characteristic meansigma methods will be calculated also by computing block 604 And newly meansigma methods is applied to output parameter 622 so that the interpolation of new feed to be described.Then, based on this new meansigma methods, corresponding to giving The feedstock characteristic of the feed level that material is drawn out of and mixability 612, tank property calculation block 604 determines will be extracted out in tank 400 The outlet characteristic 616 of feed.Then the outlet characteristic being computed is saved to complex function block 600.

Tank property calculation block 604 can comprise the some instructions for determining outlet characteristic 616, as follow-up formula describes 's.Generally, the value supposition value equal to the particular characteristics being passed to tank 400 of characteristic extracted out by each feed at outlet 406c (that is, extracting characteristic out).Can by computing block 604 adjust extract out characteristic with explanation caused by feed transmits, can be at each feed The mixing that interlayer occurs.The most known hybrid cytokine m, extracts characteristic P, and meansigma methods A of tank internal characteristic out, then exports the value of characteristic Q can be described as:

Q=(1-m) P+mA

(formula 1)

The hybrid cytokine limited during wherein m is block 612, it describes the journey of the one or more characteristics mixing in whole tank Degree, P is the value not having mixed characteristic when it is delivered to tank 400, and A is the tank recorded when feed is passed to tank The meansigma methods of the characteristic in 400.Such as tank 400 is configured to contain the slug flow of three fluid layers: comprise that pH is 4.2 first (bottom) layer, comprise the second (middle) layer that pH is 4.6 and comprise the 3rd (top) layer that pH is 4.1.Slug flow tank is configured It is 0.2 for extracting the hybrid cytokine of feed and tank out from ground floor.When the average pH of tank is 4.3, the outlet characteristic value of pH will wait In (1-0.2) (4.2)+(0.2) (4.3) or 4.22.

In view of formula 1, if being measured (that is, Q by experiment_lab) determine the value of Q, and value P of known extraction characteristic and tank Meansigma methods A of interior characteristic, it is determined that hybrid cytokine m is possible.Although directly hybrid cytokine 612 (m) can be obtained by formula 1 Optimum, but the method for least square that hybrid cytokine 612 can also be limited by the slope of the regression line most preferably mated determines, its It is described as:

Q-P=(A-P) m

(formula 2)

Determined that hybrid cytokine 612 (m) can be described as " method of least square " by the slope of the regression line, because discrete figure is The slope of the regression line of good coupling is:

(Q-P)vs.(A-P)

(formula 3)

Determined that hybrid cytokine is useful by the slope of the regression line, because not relying on shown in usual regression line formula Constant skew:

$m=\frac{\mathrm{\Σ}(x-\stackrel{\‾}{x})(y-\stackrel{\‾}{y})}{\mathrm{\Σ}{(x-\stackrel{\‾}{x})}^2}$

(formula 4)

Wherein y andAlso have x andComprise the skew of about the same amount, thereforeWithIn skew generally may be used Offset.The slope described for formula 4 calculates, y=(Q-P) and x=(A-P).

In another embodiment, calculate outlet characteristic 616 and think each characteristic P_iLimit weighted averageStart, it is assumed that Good mixing:

$\stackrel{\‾}{P_i}=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{P_i^k{w}^k}{W}$

(formula 5)

WhereinRepresent the characteristic value P of layer k_i；w^kIt it is the amount of material (weight) in layer k；W has n loaded layer The weight of the total material in tank, wherein:

$W=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{w}^k$

(formula 6)

Tank outlet characteristicCan be defined as:

$p_i^{\mathrm{outler}}=(1-m)p_i^k+m{\stackrel{\‾}{P}}_i$

(formula 7)

Wherein k=1 is hybrid cytokine corresponding to bottom loaded and k=n corresponding to top loading, l≤k≤n, and m, Wherein 0≤m≤1.As m=1, have and be thoroughly mixed, andAnd as m=0, not mixing, andI.e. export material characteristic is identical with the current primary characteristic extracted out in layer.When canned carrier has inapparent district Time other, by formula 8 this enforcement can be described:

$\stackrel{\‾}{P_i}=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{P_i^k}{n}$

(formula 8)

Fig. 7 A shows an example of the data 700 for discretization scheme 750 (Fig. 7 B).Data for chart 750 700 can comprise feedstock characteristic value Q derived by laboratory_lab702, the meansigma methods of the characteristic in extraction value P 704 of characteristic, tank A 706, the value 708 of (Q-P) and the value 710 of (A-P).Fig. 7 B shows the pass of the value 708 and value 710 of (A-P) comprising (Q-P) It it is the discretization scheme 750 of the regression line 752 of curve chart and optimal coupling.Use formula 4, usual module 115 and concrete being combined The tank property calculation block 604 of functional device 600 can comprise one or more instruction to calculate the slope of the regression line 752 of optimal coupling 712 (Fig. 7 A), i.e. hybrid cytokine m.

A kind of realization of tank property calculation block 604 comprises to give an order:

Calculating for authentication module 115 and optimization hybrid cytokine 612 exports characteristic 616, when m=0 (that is, feed to determine Layer is without mixing) and when being the slug flow not having previous layer in supposing tank 400, outlet characteristic 616 should reflect discrepancy characteristic 614.But, when m=1 (that is, being thoroughly mixed to the bed of material), outlet characteristic 616 should generally defer to the average characteristics in tank 400 Value.Fig. 8 shows exemplary plot 800, that reflects the module 115 comprising aforementioned one or more formula, and assume mixing because of Son is zero (that is, layer is without mixing) and the simple layer level in tank 400.As it can be seen, for outlet and entrance characteristic " P5 " 802, in time interval 804, when module 115 is correctly configured with formula described here, the value 806 of outlet characteristic will Defer to the given value of entrance characteristic as shown in Trendline 808.Alternatively, when there is multiple layer and hybrid cytokine in tank 400 When being arranged to 1 (that is, being thoroughly mixed), Trendline 808 is by meansigma methods (that is, the average characteristics output ginseng of the characteristic in reflection tank Several 622).Additionally, when in tank 400, only one of which layer and hybrid cytokine are 1, the figure of module 115 checking will comprise with in Fig. 8 The Trendline 808 (that is, the value 806 of outlet characteristic defers to the value of known entrance characteristic) that the line represented is similar.

Alternatively, modelling verification can comprise the calculating of outlet characteristic Q of several values for hybrid cytokine m, and can With the overall non-match error using the experimental result of value of Q to obtain, it is each that the value of this Q can illustrate for hybrid cytokine 612 The outlet characteristic 502 of value.By using the method, the hybrid cytokine producing minimal overall error will be selected for module 115 In.The stator that is limited optimizes hybrid cytokine, it is possible to use several blend factor values (such as, 0.0,0.25,0.5,0.75 and 1.0)。

With reference to formula 1, also can be adjusted value P extracting characteristic out by hybrid cytokine m, to illustrate that multiple layer is for extracting layer out (i.e., Current the most just by extraction from tank 400 in the bed of material) proximity.Such as, the feed of tank it is passed to by top entry 404a The impact of the bottom layer of comparison tank is had more by the impact on the top layers of tank, and vice versa.Diffusion, convection current and feed Other factors of movement and composition influence hybrid cytokine and a layer are inversely proportional to interfloor distance to the mixed influence of another layer.Can Think that two or more layers calculate correction.For multilamellar correction, application depends on the distance of loaded layer, the decline letter of mixing Number.Can apply by using linear, index or other functions.The characteristic being corrected of new loaded layer can be retouched by formula 9 State:

$p_i^1\left(\mathrm{corr}\right)=(1-m_1)p_i^1+m_1{p}_i^2$

(formula 9)

In a similar fashion, the carrying out that the correction of adjacent layer can be described by formula 10 calculates:

$p_i^2\left(\mathrm{corr}\right)=(1-m_1)p_i^2+m_1{p}_i^1$

(formula 10)

Multilamellar correction with two-layer correction is sequentially applied to layer right: layer 1-2,2-3,3-4 etc. are similar.Can arbitrarily set warp The number of the layer revised, if fewer than the material layer number in tank when loading.

Multilamellar depends on layer to the impact extracting layer out and reduces position in a linear fashion, can be described by formula 11:

$m_i=m_1\frac{l-i+1}{l}$

(formula 11)

Wherein l represents the quantity of layer having been for being corrected before the impact of other layers, and i is the layer i, i+1 that application is revised Index, i≤l, i≤n, n are the exact number of the layer in tank with feed, and m₁Revising hybrid cytokine, it is defined as The part of hybrid cytokine m (assuming that m₁=(0.1 to 0.5) m).To be revised if plurality of layers comprises, the index that formula 12 describes Revise and the mixing of two adjacent layers will be able to be more preferably described:

$m_i=m_1{e}^{\frac{i-1}{l}}$

(formula 12)

Layer revises hybrid cytokine m₁To be applied to calculate the hybrid cytokine m of extraction characteristic less than (such as, 2 to 5 times).This Assuming that a reason be m₁Experimental verification highly difficult and correction may be crossed.Can also arbitrarily set and there is the spy being corrected Quantity l (such as, being defaulted as 2 to 4 layers) of the layer of property.For m₁Can allow to save with l selection fixed value and determine these factors Additional step.Alternatively, the method can apply a proving program only to limit final hybrid cytokine, this and do not repair having The operation that the archetype of positive layer is carried out is identical.Property calculation in formula 9 and 10 can be assumed that adjacent layer has identical weight. Different layers weight can introduce additional calculating error.

Fig. 9 is can be performed to realize above-mentioned feedstock characteristic MBM 115 and the example of tank property calculation block 604 The flow chart of property method.This flow chart can basically describe for determining without the feedstock characteristic in the storage tank of blender Method 900.Although the method 900 is described as including the configuration of slug flow type inlet/outlet, but the method 900 is all right It is used to support that other kinds of tank as described herein configures.The method 900 can include one or more can with computer Perform instruction for the function of form or routine, these one or more functions or routine be stored in computer-readable memory and The processor using calculating equipment performs.These routines can be included as the part of feedstock characteristic MBM 115 (Figure 1B).

In function 902, module 115 may determine that new feed has been passed to storage tank 400 (Fig. 4) the most.One In a little embodiments, module 115 can include the instruction inquiring about stack 500 (Fig. 5), to determine whether time response 508 includes and close Join the value different from the value of date-time block 620 (Fig. 6 A and 6B).Such as, when when being newly delivered to reach storage tank 400 of feed, use Characteristic 502 in this new feed can be pressed in stack 500, including time value 508.Can be by new feed time value 508 Compare with current time of day external input parameter 620.New timestamp in stack 500 may indicate that have new characteristic 502 The transmission of new feed 420.

If the transmission of feed 402 includes new timestamp, then function 904 can use stack 500 and complex function block 600 perform some calculating.In certain embodiments, function 904 can update stack 500, and is current total in storage tank 400 Each characteristic 502 of body feed 402 calculates meansigma methods.The calculating performed by function 904 can be based on storage tank level 608 and stack Value 504 in 500.If this transmission does not include new timestamp, then can be with skip functions 904.

In function 906, module 115 may determine that the position of the new feed in tank 400.In certain embodiments, newly give Material can be passed to the various positions in tank 400.Such as, new feed can be passed to the top 404a of tank 400, middle part 404b or bottom 404c.In the tank 400 of slug flow configuration, permissible in the transmission of top 404a or the feed of middle part 404b The layering of the feed 402 in index pot 400 and new to the bed of material creating to the top of the bed of material before.In function 908, The newest feed is passed to top or the middle part of tank 400, and the value of the characteristic of the feed then extracted out from tank 500 is permissible Level 608 based on the feed in tank.Such as, with reference to Fig. 4, exemplary tank 400 has existing feed, and quilt at layer 408 It is configured to the slug flow with the outlet at position 406c.New feed transmission can enter tank 400 at entrance 404a or 404b And create new to the bed of material 410.In this illustration, due to transmission time currently have in tank before one to the bed of material (408), will by from tank 400 extract out feed by based on before that in tank to the bed of material (or level) or have Some between this front layer 408 and new layer 410 mix to the bed of material 408.Therefore, function 908 can be by outlet characteristic 616 It is set to off being coupled to before to the characteristic of the bed of material (that is, to the bed of material 408).But, it is passed to slug flow configuration at new feed Tank bottom in the case of, there is not or only occur very small amount of mixing, and new feed will Directly extracted out from tank.Therefore, in the case of new feed is passed to the pot bottom of slug flow configuration, function 910 can Outlet characteristic 616 to be set to the characteristic of the feed 402 newly transmitted.

In function 912, module can be based on hybrid cytokine 612 and the feedstock characteristic of current feed in tank 400 Outlet parameter 622 calculate outlet characteristic 616.In certain embodiments, during function 912 can use formula described herein One or more based on hybrid cytokine 612 and the average characteristics value of current feed in tank to calculate new outlet special Property 616.

In function 914, module 115 can store the outlet characteristic 616 by using function 912 to calculate.Real at some Executing in example, outlet characteristic 616 can be stored in complex function block 600.

Feed leave storage tank 400 and export characteristic 616 by calculate after, outlet feed can enter other mistake Journey plant entities, such as, reactor 407.It is current at reactor that other module 410 can use outlet characteristic 616 to calculate The various characteristics of the feed in 407.Such as, when when new feed arrives, old feed removes, outlet characteristic 616 can be in time Change.Therefore, module 410 can follow the tracks of and record the characteristic of the feed entering reactor 407, with true on any preset time Determine the composition of feed in reactor 407.

The illustrative methods 900 of Fig. 9 can be performed by processor, controller and/or arbitrarily other processing equipments being suitable for. Such as, method 900 can be embodied in the coded command being stored in the most tangible computer-readable medium, and this is tangible Computer-readable medium such as flash memory, CD, DVD, floppy disk, ROM, RAM, programming ROM (PROM), electrically programmable ROM (EPROM), electric erasable PROM (EEPROM), optical memory disc, light storage device, magnetic storage disk, magnetic storage apparatus and/or Arbitrarily other can be used the program coding and/or the medium of instruction carrying or storing with method or data structure as form, And its can by processor, universal or special computer or other there is the machine of processor (such as, below in conjunction with Figure 10 The example processor platform P10 discussed) use.Combination above is also included within the range of computer-readable medium.

Method includes, such as so that processor, general purpose computer, special-purpose computer or dedicated processes machine realize one Individual or the instruction of multiple ad hoc approach and/or data.Alternatively, it is possible to use ASIC, PLD, FPLD, discreet logic, hardware, solid The most one or more combination of part etc. carrys out some or all of implementation method 900.

And it is possible to alternatively employing manual operation realizes some or all of illustrative methods 900, or it is permissible It is embodied as any combination of any aforementioned techniques, such as firmware, software, separation logic and/or the combination in any of hardware.Additionally, Many other can be used to realize the method for exemplary operation of Fig. 9.Such as, the execution sequence of function can be modified, and/or One or more described functions can be modified, deletes, Further Division or combination.Additionally, any or all of show Example method 900 can be sequentially performed and/or by such as, discrete process thread, processor, equipment, separation logic, circuit Etc. being performed in parallel.

Illustrative methods 900 characteristic based on processor control system is special to that measure, that calculate and/or whole feed Relationship modeling between property.Multiple illustrative methods 900 can be performed in parallel or in serial with to Process Control System Part modeling and/or other Process Control Systems are modeled.

Figure 10 is the module that can be used for realizing above-mentioned illustrative methods with example processor system P10 of equipment Figure.Such as, similar or like with example processor system P10 processor system can be used for realizing Fig. 1 and/or 4 In exemplary OMS 102, exemplary lot data receptor 402, example analytic processor 114, exemplary analysis process Process model building device 408, exemplary assessment process model building device 410, example process modeling maker 412, exemplary display management In device 420, exemplary session controller 422, exemplary online data processor 116, exemplary factory access server 424, And/or exemplary network accesses server 428.Although example processor system P10 be described as later including multiple outside Portion's equipment, interface, chip, memorizer etc., these one or more elements can be by from for realizing exemplary OMS 102, exemplary lot data receptor 402, example analytic processor 114, exemplary analysis processing procedure modeling device 408, Exemplary assessment process model building device 410, example process modeling maker 412, example display manager 420, exemplary meeting Words controller 422, exemplary online data processor 116, exemplary factory access server 424 and/or exemplary network Access in other the one or more example processor system in server 428 and omit.

As shown in Figure 10, processor system P10 includes the processor P12 being couple to interconnection bus P14.Processor P12 bag Including register set or register space P16, it is all on chip as shown in Figure 10, but its can alternatively part or It is coupled directly to processor the most not on chip and by special electrical connection and/or by interconnection bus P14 P12.Processor P12 can be processor, processing unit or the microprocessor being arbitrarily suitable for.Although being not shown in Figure 10, it is System P10 can be multicomputer system, and therefore can include one or more additional processor, this additional processor Similar with processor P12 or identical, and it is communicatively coupled to interconnection bus P14.

The processor P12 of Figure 10 is coupled to chipset P18, and this chipset P18 includes memory manager P20 and outer Enclose input/output (I/O) controller P22.As is well known, the commonly provided I/O of chipset and memory management functions, with And multiple general and/or special register, intervalometer etc., it can be visited by one or more processors being couple to chipset P18 Ask or use.Memory Controller P20 performs to make processor P12, and (or multiple processor, if there being multiple processor Words) it is able to access that system storage P24 and the function of mass storage P25.

System storage P24 can include the volatilization of any desirable type and/or nonvolatile memory such as, static random Memorizer (SRAM), dynamic RAM (DRAM), flash memory, read only memory (ROM), etc..Mass storage P25 can include the mass-memory unit of any desirable type.Such as, if example processor system P10 is used for reality Existing OMS 102 (Fig. 2), then mass storage P25 can include hard disk drive, CD drive, fc tape storage device FC, Deng.Alternatively, if example processor system P10 is used for realizing process model data base 416 and/or lot data processes Device 406, then mass storage P25 can include solid-state memory (such as, flash memory, RAM memory etc.), magnetic storage Device (such as, hard disk drive) or be suitable for the great Rong in process model building data base 416 and/or lot data data base 406 Other memorizeies any of amount storage.

Peripheral I/O controller P22 perform to make processor P12 can by peripheral I/O bus P32 with peripheral input/defeated Go out the function that (I/O) equipment P26 with P28 and network interface P30 communicates.I/O equipment P26 with P28 can be any type of I/O equipment such as, keyboard, display (such as, liquid crystal display (LCD), cathode ray tube (CRT) display, etc.), navigator (example As, mouse, trace ball, capacitive type touch pad, handle etc.), etc..Network interface P30 is it may be that such as ethernet device, asynchronous Transmission mode (ATM) equipment, 802.11 equipment, DSL modem, wire line MODEM, mobile telephone modem Deng, it makes the processor system P10 can be with other processor system communication.

Although Memory Controller P20 and I/O controller P22 is the functional device separated in chipset P18 as shown in Figure 10, The function performed by these blocks can be integrated in single semiconductor circuit, or two or more can be used to separate Integrated single channel realizes.

At least some of above-mentioned illustrative methods and/or device is one or more soft by run on a computer processor Part and/or firmware program realize.But, include but not limited to that special IC, programmable logic array and other hardware set Standby specialized hardware realizes can being built as equally wholly or partly realizing illustrative methods described herein and/or dress Some or all put.Additionally, include but not limited to distributed treatment or component/object distributed treatment, parallel processing or void The software of the replacement that plan machine processes realizes also being able to be built as realizing illustrative methods described herein and/or system.

It should be noted that, example software described herein and/or firmware realize being stored on tangible media, such as: Magnetic medium (such as, disk or tape)；Magneto-optic or optical medium such as CD；Or solid state medium, such as storage card or receiving Other of one or more read-only (non-volatile) memorizer, random access memory or other rewritable (volatibility) memorizeies Encapsulation.Therefore, example software described herein and/or firmware can be stored in tangible media and such as above or follow-up retouch On the storage medium stated.Superincumbent description quotes specific standard and agreement to describe the degree of example components and function On, it should be appreciated that the scope of this patent is not limited to such standard and agreement.Such as, for the Internet and other packet switching networks Network transmission (such as, transmission control protocol (TCP)/Internet protocol (IP), UDP (UDP)/IP, hypertext mark Note language (HTML), HTML (Hypertext Markup Language) (HTTP)) each standard represent the example of this area current state.Such mark Quasi-periodically by have identical general utility functions, faster replace with more effectively equivalents.Therefore, there is identical function Replacement standard and agreement be contemplated by this patent, and be intended to the equivalent shape included within the scope of the appended claims Formula.

Although illustrative methods described below and equipment include, perform in other assemblies on hardware, software and/or Firmware, it should be noted that these examples are merely illustrative, and it is not considered as restrictive.For example, it is envisioned that arbitrary or All of hardware, software and fastener components can uniquely with hardware, uniquely with software or with the combination of hardware Yu software Embody, therefore, although described below illustrative methods and equipment, this area commonly calculates and is arbitrarily appreciated that these are provided Example be not to realize the sole mode of this method and apparatus.

Although being included on hardware, additionally, this patent discloses, the software or the illustrative methods of firmware and device performed, It should be noted that such system is merely illustrative, and be not considered as restrictive.For example, it is envisioned that these hardware and soft In part parts any one or all can uniquely with hardware, uniquely with software, uniquely with firmware or with hardware, firmware And/or certain embodied in combination of software.Therefore, although description above describes illustrative methods, system and machine and can visit Asking medium, these examples are not the sole modes realizing such system, method and machine accessible medium.Therefore, although There is described herein some illustrative methods, system and machine accessible medium, the coverage of this patent is not limited to this.

Claims (25)

1. a method for the feed material characteristic in the storage tank without blender determining process control plant, described side Method includes:

It is applied to previous newly to the bed of material to the feedstock characteristic value of the bed of material；

Extraction layer is set up based on the storage tank horizontal survey being associated with tank outlet；

Calculate the meansigma methods of the feedstock characteristic of total feeding coal in described storage tank；

Calculate the hybrid cytokine of described extraction layer；And

Meansigma methods based on described feedstock characteristic and described hybrid cytokine calculate the extraction feedstock characteristic value of described extraction layer；

Feed in wherein said extraction layer mixes with other feed section in described storage tank.

2. the method for claim 1, it is characterised in that build based on the storage tank horizontal survey being associated with tank outlet Vertical layer of extracting out includes the renewal stack physical state with the described feed in the described storage tank of reflection.

3. the method for claim 1, it is characterised in that the outlet of described tank includes slug flow.

4. method as claimed in claim 3, it is characterised in that build based on the storage tank horizontal survey being associated with tank outlet The vertical described new position to the bed of material extracted out in layer includes determining described storage tank.

5. method as claimed in claim 4, it is characterised in that be newly positioned at tip position or interposition to the bed of material if described Putting, the described extraction feedstock characteristic value of the most described extraction layer is by level based on the total feeding coal in described storage tank.

6. method as claimed in claim 4, it is characterised in that be newly positioned at bottom position to the bed of material if described, then described in take out The described extraction feedstock characteristic value going out layer comprises the described new described feedstock characteristic value to the bed of material.

7. the method for claim 1, it is characterised in that the extraction feedstock characteristic value farther including to be computed updates Data structure is with the physical state of the described feed in the described storage tank of reflection.

8. the method for claim 1, it is characterised in that described hybrid cytokine comprises the recurrence of the optimal coupling of discrete figure The slope of line, wherein, described discrete figure indicates feedstock characteristic value Q drawn by experiment_labDeduct the feedstock characteristic value without mixing The value that P obtains and the ratio being deducted the value that described feedstock characteristic value P without mixing obtains by meansigma methods A of feedstock characteristic.

9. the method for claim 1, it is characterised in that calculate extraction feedstock characteristic value Q and include:

Q=(1-m) P+mA

Wherein m is described hybrid cytokine, and P is the feedstock characteristic value without mixing, and A is the meansigma methods of described feedstock characteristic.

10. the method extracting feed material characteristic out in the storage tank without blender calculating batch processing, described side Method includes:

Calculate the meansigma methods of the feedstock characteristic of total feeding coal in storage tank, described total feeding coal comprise new in the bed of material to Material；

Determining the described new position to the bed of material in described tank, described position is corresponding to the entry position of described storage tank；

Determining the position extracting layer out in described tank, described extraction layer is corresponding to the exit position of described storage tank；

Based on feedstock characteristic value Q drawn by experiment_labDeduct the value of the feedstock characteristic value P acquisition without mixing and by feedstock characteristic Meansigma methods A deduct the regression line of optimal coupling of discrete figure of ratio of the value that described feedstock characteristic value P without mixing obtains Slope calculate extraction feedstock characteristic value.

11. methods as claimed in claim 10, it is characterised in that described storage tank is slug flow storage tank.

12. methods as claimed in claim 10, it is characterised in that described entry position include tip position, centre position and In bottom position one.

13. methods as claimed in claim 10, it is characterised in that described exit position includes tip position, centre position and In bottom position one.

14. methods as claimed in claim 10, it is characterised in that calculate extraction feedstock characteristic value Q and include:

Q=(1-m) P+mA

Wherein m is described slope, and P is the described feedstock characteristic value without mixing, and A is the meansigma methods of described feedstock characteristic.

15. methods as claimed in claim 10, it is characterised in that described feedstock characteristic value comprises PH balance, degree of reaction, poison One or more in property, concentration, density, molecular weight and viscosity.

16. methods as claimed in claim 10, it is characterised in that slope value be 0 corresponding to entirely without mixing feedstock characteristic, And slope value is 1 corresponding to the feedstock characteristic being thoroughly mixed.

17. methods as claimed in claim 10, it is characterised in that in described tank multiple other to the bed of material to described extraction layer Impact comprise the hybrid cytokine m being indexed as layer i and extracting out layer_iIndex correction:

$m_i=m_1{e}^{\frac{i-1}{l}}$

Wherein l be previously be corrected due to the impact of other layers multiple to the quantity of the bed of material, i is by the described layer affected Index, i≤l, i≤n, n be in described storage tank existing, other are to the quantity of the bed of material；And m₁It is to revise hybrid cytokine.

18. 1 kinds of feeds in the storage tank without blender determining process control plant implemented in computer installation The method of substance characteristics, described method comprises the steps of

I. the new feed transmission of storage tank is detected；

II. with the feedstock characteristic value more new data structure transmitted corresponding to described new feed and calculate in described storage tank always to The meansigma methods of the feedstock characteristic of doses, described total feeding coal comprises new to the new feed transmission in the bed of material；

III. determine the described new position to the bed of material in described tank and extract the position of layer out；And

IV. hybrid cytokine based on described extraction layer calculates extraction feedstock characteristic value Q, and described hybrid cytokine comprises by testing Feedstock characteristic value Q gone out_labDeduct the value of the feedstock characteristic value P acquisition without mixing and deducted by meansigma methods A of feedstock characteristic described The slope of the regression line of the optimal coupling of the discrete figure of the ratio of the value that feedstock characteristic value P of nothing mixing obtains.

19. methods as claimed in claim 18, it is characterised in that also comprise renewal stack in described step II to reflect described storage The physical state of the described feed in batch can.

20. methods as claimed in claim 18, it is characterised in that be newly positioned at tip position or interposition to the bed of material if described Put, then calculate described extraction layer, the extraction feedstock characteristic value of level based on the total feed in described storage tank.

21. methods as claimed in claim 18, it is characterised in that be newly positioned at bottom position to the bed of material if described, then calculate The extraction feedstock characteristic value of described extraction layer, described extraction layer extract out feedstock characteristic value comprise described new to the bed of material described to Material characteristic value.

22. methods as claimed in claim 18, it is characterised in that described storage tank is slug flow storage tank.

23. methods as claimed in claim 18, it is characterised in that calculate described extraction feedstock characteristic value Q and include:

Q=(1-m) P+mA

Wherein m is described slope, and P is the described feedstock characteristic value without mixing, and A is the meansigma methods of described feedstock characteristic.

24. methods as claimed in claim 18, it is characterised in that described feedstock characteristic value comprises PH balance, degree of reaction, poison One or more in property, concentration, density, molecular weight and viscosity.

25. methods as claimed in claim 18, it is characterised in that slope value be 0 corresponding to entirely without mixing feedstock characteristic And slope value is 1 corresponding to the feedstock characteristic being thoroughly mixed.