CN101957597B - Real-time optimizer in continuous production process - Google Patents
Real-time optimizer in continuous production process Download PDFInfo
- Publication number
- CN101957597B CN101957597B CN2010102895041A CN201010289504A CN101957597B CN 101957597 B CN101957597 B CN 101957597B CN 2010102895041 A CN2010102895041 A CN 2010102895041A CN 201010289504 A CN201010289504 A CN 201010289504A CN 101957597 B CN101957597 B CN 101957597B
- Authority
- CN
- China
- Prior art keywords
- time
- variable
- real
- optimization
- transfer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 230000008569 process Effects 0.000 title claims abstract description 18
- 238000010924 continuous production Methods 0.000 title abstract 2
- 238000005457 optimization Methods 0.000 claims abstract description 94
- 238000004519 manufacturing process Methods 0.000 claims abstract description 52
- 238000012545 processing Methods 0.000 claims abstract description 9
- 238000012546 transfer Methods 0.000 claims description 54
- 238000006243 chemical reaction Methods 0.000 claims description 39
- 230000008859 change Effects 0.000 claims description 34
- 239000000047 product Substances 0.000 claims description 23
- 238000013178 mathematical model Methods 0.000 claims description 20
- 238000005259 measurement Methods 0.000 claims description 16
- 239000002994 raw material Substances 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 7
- 238000000926 separation method Methods 0.000 claims description 6
- 239000007795 chemical reaction product Substances 0.000 claims description 5
- 238000012821 model calculation Methods 0.000 claims description 5
- 238000003745 diagnosis Methods 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 230000009977 dual effect Effects 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 230000002452 interceptive effect Effects 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 3
- 238000012544 monitoring process Methods 0.000 abstract 1
- 238000005336 cracking Methods 0.000 description 12
- 238000000197 pyrolysis Methods 0.000 description 11
- OFBQJSOFQDEBGM-UHFFFAOYSA-N Pentane Chemical compound CCCCC OFBQJSOFQDEBGM-UHFFFAOYSA-N 0.000 description 8
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 7
- 239000001257 hydrogen Substances 0.000 description 7
- 229910052739 hydrogen Inorganic materials 0.000 description 7
- 150000004945 aromatic hydrocarbons Chemical class 0.000 description 6
- VGGSQFUCUMXWEO-UHFFFAOYSA-N Ethene Chemical compound C=C VGGSQFUCUMXWEO-UHFFFAOYSA-N 0.000 description 4
- 238000005553 drilling Methods 0.000 description 4
- 239000007789 gas Substances 0.000 description 4
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 4
- QQONPFPTGQHPMA-UHFFFAOYSA-N propylene Natural products CC=C QQONPFPTGQHPMA-UHFFFAOYSA-N 0.000 description 4
- 125000004805 propylene group Chemical group [H]C([H])([H])C([H])([*:1])C([H])([H])[*:2] 0.000 description 4
- 238000001833 catalytic reforming Methods 0.000 description 3
- 239000000470 constituent Substances 0.000 description 3
- 239000004215 Carbon black (E152) Substances 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004523 catalytic cracking Methods 0.000 description 2
- 239000000571 coke Substances 0.000 description 2
- 238000004939 coking Methods 0.000 description 2
- 230000010429 evolutionary process Effects 0.000 description 2
- 229930195733 hydrocarbon Natural products 0.000 description 2
- 150000002430 hydrocarbons Chemical class 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 239000003208 petroleum Substances 0.000 description 2
- 239000005977 Ethylene Substances 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000003556 assay Methods 0.000 description 1
- 238000003339 best practice Methods 0.000 description 1
- 239000003054 catalyst Substances 0.000 description 1
- 238000011208 chromatographic data Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 239000002283 diesel fuel Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000003209 petroleum derivative Substances 0.000 description 1
- 239000000376 reactant Substances 0.000 description 1
- 238000002407 reforming Methods 0.000 description 1
- 238000006057 reforming reaction Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Images
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention relates to a real-time optimizer in a continuous production process. The real-time optimizer consists of a production process monitor (1), an optimization object real-time calculator (2), an intelligent optimizing controller (3), an optimization controller (4) and an operation monitoring and setting device (5). The real-time optimizer acquires operating data of a production process (7) in real time through an existing data input/output interface (6), and sends the optimization function to the production process (7) to perform optimization; and in the optimization of the practical production process, the optimization means can quickly and accurately adjust the optimization variable, maintain the optimal operating condition under various interferences and perform fault detection and processing of the production process and detection and processing of various constraints.
Description
Technical field
The present invention relates to a kind of real-time optimization device that makes production run operate in the continuous flow procedure under the optimization state at any time.
Background technology
Many to continuous flow procedure; The size of its optimized operation situation and optimization target values (following all be up to optimum condition with desired value) all receives to survey the influence of variable and often changes; Optimization target values also can not be surveyed under many situation; How making production run operate in optimum condition at any time, is the problem that real-time optimization will solve.
For solving the real-time optimization problem; To many continuous flow procedures; Mostly adopt method: common chemical reaction process in the petrochemical production process for example based on the production process mathematics model of stable state; When operation; Raw material is formed (can not survey) and is often changed, and operating condition and running environment (wherein some variable can not be surveyed) also often change; Optimization aim, like the conversion ratio and the productive rate of reaction, or selectivity and reactant composition, also usually can not survey.For reactor is operated under the optimum condition (the peak optimization reaction degree of depth); Method commonly used is to form and catalyst activity according to the raw material that assay obtains; Actual measurement process variable when production run is in stable state is calculated the optimum operating conditions of target of sening as an envoy to mathematics model of stable state, and course of reaction is adjusted; The excellent variable of modal accent is a temperature of reaction, promptly adjusts temperature of reaction to reach optimum.Each transfer to wait for that production process operates steadily after excellent after, transfer excellently once more based on the actual measurement variable of production process and model, twice excellent interval of accent (transferring the excellent cycle) is generally more than two hours.Above method weak point is: the one, and real-time is poor, can not when production run changes, adjust to rapidly under the optimization situation and move.The 2nd, can not adapt to the dynamic change of production run actual motion; The actual measurement process variable of institute's foundation is always devious, adds that the precision of model is limited, makes to calculate gained optimal conditions and actual can not coincideing; Therefore, also seldom see effective instance of this method of foundation.
Another kind of optimization method is without model, only how to transfer next time excellently according to the variation decision of variation of transferring excellent variable and optimization aim, and it transfers excellent logic to be: when desired value increases when identical with the excellent increment change direction of accent, continue to increase the excellent variable of accent.(cycle) accent is once excellent at regular intervals, makes production run progressively approach the optimized operation situation, abbreviates as from seeking best practice. and the advantage of this method is simple, has avoided the inaccurate influence of model; Owing to, do not require the measured value absolutely accurate according to target and the relative variation of transferring excellent variable yet; Its main weak point is: the one, desired value and the change direction of transferring excellent variable are judged, if incompatibility production run dynamic perfromance possibly cause erroneous judgement.The 2nd, near the optimized operation situation, possibly transfer excellent direction all changing at every turn, form the vibration of transferring excellent variable, make the production run can not even running.The 3rd, when desired value or the variable relevant with desired value do not have real measured data,, compare with above-mentioned method with Model Calculation like the conversion ratio and the selectivity of above-mentioned chemical reactor, just can't implement to optimize.
When reality is implemented production process optimization, also need the excellent means of good accent, can adjust excellent variable rapidly accurately, can under various interference, keep the optimized operation situation again.The fault detect of production run and processing, the detection of various constraints also is necessary with handling.
Summary of the invention
The objective of the invention is to provide one and adapt to production run operation dynamic perfromance and actual conditions, effective real-time optimization device.
The real-time optimization device of continuous flow procedure of the present invention, its principal character is:
The real-time calculating of optimization aim: when optimization aim can not be surveyed, according to measurable process variable and dynamic mathematical models, online in real time calculated.
" time-out " of optimizing and " self-starting ": after optimizing makes production process optimization, or the production run situation promptly suspends optimizing when not allowing, and avoids near the vibration optimized point, prevents the abnormal running situation, remains on even running under the optimization situation.
Need to transfer when excellent when production run and environment thereof change, start optimizing automatically, make production run be in the optimization situation at any time.
" dynamic compensation ": according to the dynamic perfromance of producing with searching process, optimization target values in the optimizing cycle is implemented dynamic compensation, can optimizing when the production run dynamic change, avoid the erroneous judgement of optimizing, shorten the optimizing cycle, acceleration optimizing process.
Constraint is forecast and handled: based on dynamic mathematical models variable is tied and forecasts, when forecast was tied (reaching bound), the direction that only allows court to break away from constraint was transferred excellent, and optimizing is moved under safe and reliable condition.
Data processing and fault diagnosis: to the process variable that contains noise with optimization target values is handled and fault diagnosis, optimizing is being carried out reliably on the basis.
Intelligence optimizing: on the whether identical basic optimizing logical foundations of the change direction of judge transferring excellent variable and optimization aim, increase judge intelligent, in time " time-out " or " self-starting ", the shortening searching process is guaranteed optimizing safety.
Based on the model pre-estimating coordinating control system of dynamic model as transferring excellent means: provide rapid response, can steadily keep the excellent means of accent of optimization situation again.
The real-time optimization device of continuous flow procedure of the present invention is by production run monitor 1, the real-time counter 2 of optimization aim, and intelligent optimizing controller 3 is transferred excellent controller 4, and operation is kept watch on setting apparatus 5 five parts and is formed.The real-time optimization device is obtained the service data of production run 7 in real time through existing data input/output interface 6, and will transfer excellent effect to give production run 7 and transfer excellently, forms the real-time optimization system.
1. production run monitor: mainly contain following function:
1-1. data processing and fault diagnosis: most of prior art capable of using.A characteristic of this patent is based on the poor of dynamic mathematical models result of calculation and actual measurement variable, and whether real-time judge operating condition and real measured data be normal.
1-2. constraint forecast: according to dynamic model following variation of variable estimated, realized whether the be tied forecast of (reaching bound) of variable, forecast stop when transfiniting the transfiniting accent of direction is excellent." simplification " and " complete " dual mode is arranged having estimated of variable is following:
Simplified way: Y (t+p/t)=Y (t)+A [Y (t)-Y (t-1)] (1)
The t=current time, p=estimates time domain, the prediction coefficient that A=can be provided with
Y (t+p/t)=current time is to following t+p variable discreet value constantly
Y (t), the measured value of Y (t-1)=current time and previous moment variable
Complete mode: based on the discrete time state-space model,
Model: X (k+1)=F [X (k-τ
X), U (k-τ
U)] (2A)
Y(k)＝G[X(k)，U(k)] (2B)
Y=need estimate following variable X=state variable U=input variable
K=t/T
d=discrete time periodicity t=time T
d=discrete time the cycle
τ
XBetween=the state variable interactive retardation time corresponding discrete time periodicity
τ
U=input variable is to the discrete time periodicity of the retardation time of state variable influence
Forecast through the online in real time correction:
Y
C(k+p/k)＝Y(k+p/k)+[Y(k)-Y(k/k-p)] (3)
Y
C(k+p/k)=revised [at following (k+p) T
dPredicted value constantly]
P is the corresponding discrete time periodicity of forecast time institute of setting
Y (k+p/k)=by (k+p) T in future of Model Calculation
dDiscreet value constantly
Y (k/k-p)=by the current time (kT of Model Calculation
d) discreet value
The measured value of Y (k)=current time k,
Y (k)-Y (k/k-p)=online in real time (to model pre-estimating) modified value.
With revised predicted value Y
C(k+p/k) whether judgment variable transfinites.
1-3. working conditions change detects: normal in the production run operation, no variable is tied, under the trouble-free situation; Owing to can not survey the variation of variable; When making production run operation depart from the optimization situation, this patent detects this working conditions change in real time, for the self-starting of optimizing provides information.
Working conditions change provided by the invention detects content and comprises: whether the size and the direction of each variable measured value have significant change when each variable that 1. relatively calculates based on state-space model [formula (2A) (2B)] and actual motion.2. the technological parameter that calculates based on the dynamic mathematical models online in real time has significant change, like the heat and mass transport coefficient, and concentration of component, etc.3. optimization target values has significant change.4. operating condition has significant change, like self-actuated controller or artificial performance variable significant change is arranged, and production decision changes, and raw material is formed and fluctuations in discharge, the variable bound changed condition, and production equipment changes, or the like.
2. the real-time counter of optimization aim: for the optimization aim that can not survey; The present invention is according to production run mechanism; The actual measurement variable; Dynamic model [formula (2A) (2B)] and the time lag that exists, online in real time is calculated and is provided the optimization aim that can not survey, when for example the conversion ratio of chemical reactor or the product in the reaction product are distributed as optimization aim; The measured discharge of each product that can be separated by its downstream product separation equipment and the dynamic mathematical models of separation equipment are calculated the flow of separation equipment inlet (being reactor outlet) each product, thereby calculate corresponding optimization aim.During calculating, all to carry out appropriate Filtering Processing, to suppress noise to actual measurement variable and result of calculation.
Another characteristics of the real-time counter of optimization aim are: when any the actual measurement variable that is used to calculate has fault or when improper, all provides failure message, optimizing is suspended.
3. intelligent optimizing device: realizing the time-out and the self-starting of optimizing, the variable quantity of the excellent back of each accent desired value is carried out dynamic compensation, stop to transfer excellent when certain condition or operating mode do not allow satisfying, is the major function of intelligent optimizing device.
Characteristics of intelligence optimizing device are " self-starting " optimizing: after not having constraint and possessing the optimizing data of enough time spans, the operator provides and transfers excellent order, or the operating mode judgement needs to transfer excellent; Or constraint release; Or (transferred last time excellent after) excellent time out of accent of setting arrives, and the optimizing device will start optimizing automatically, promptly according to the change direction of transferring excellent variable and optimization aim; Confirm to transfer the adjustment direction and the size of excellent variable, transfer excellent.Transfer for the first time when excellent,, can confirm the excellent direction of accent based on condition of setting and current working if transfer excellent variable or optimization aim no change.Each transfer excellent after, through the accent set after excellent cycle time, beginning for the second time and after each time accent excellent.
Second characteristic of intelligence optimizing device are to transfer the variation of optimization target values in the excellent cycle to have dynamic compensation to each; This is because following phenomenon can appear in the dynamic perfromance of production run and data processing: transfer at every turn excellent before, descend as if optimization aim, even transfer excellent effect to make on the target; Transferring what show after excellent possibly be that desired value is also descending; The speed that just descends is slack-off, and when an optimizing end cycle, optimization target values possibly not rise; If think that optimization aim does not rise, and will cause the accent of anisotropy excellent.The effect of dynamic compensation is exactly to prevent that wrong accent is excellent, thereby shortens the excellent process of transferring.Dynamic compensation utilizes dynamic model [method that can select for use (1) formula or (3) formula to provide], calculates when transferring excellent amount constant, transfers in the excellent cycle the 1st, 2 at one, up to the variation of N calculation time optimization aim, is called zero of optimization aim and responds.Use the optimization target values and zero respond poor of transferring excellent back practical manifestation to come out, as judging the foundation of transferring excellent direction.
The 3rd characteristics of intelligence optimizing device be in time stop to transfer excellent: multiple time-out standard can be set, make and transfer excellent process to stop, as: transfer excellent number of times to reach setting value continuously towards a direction; Twice accent is excellent in the opposite direction, transfers excellent back optimization aim not have significant change, transfers excellent suffering restraints; Used real measured data or online in real time are calculated has fault, and abnormal condition appears in production run, and excellent data are transferred in the judgement that does not possess sufficient length; Or the like
4. transfer excellent controller: the excellent effect of the accent that intelligent optimizer provides is implemented to transfer excellent through transferring excellent controller to production run.Employing based on the model pre-estimating coordinating controller of dynamic mathematical models as transferring excellent controller.The controller of recommending employing Chinese patent 99105546.2 and 200710118864.3 to provide based on model; Optimization for chemical reactor; What recommendation employing Chinese patent 200410048083.8 provided is the automatic control system of controlled variable with macroreaction heat, as transferring excellent controller.
5. operation is kept watch on and setting apparatus: the operation of real-time optimization device is had only a run switch.Can keep watch on and transfer excellent operation conditions, the function and the operational factor of real-time optimization device are carried out online adjustment.
Description of drawings
Fig. 1 is the composition frame chart of the real-time optimization device that provides of the present invention.
Embodiment
Embodiment 1: the real-time optimization of drilling process
Optimization aim is that drill speed is the highest.In general, the pressure of the drill is high more, and ROP is high more, but the rate of wear of drill bit is also high, therefore, has the pressure of the drill of an optimum.ROP can be measured in real time, but possible ROP depends on the character of the rock stratum of boring, the degree of wear of drill bit; The flow velocity of drilling fluid and character, or the like, these factors that influence drill speed all do not have real-time metrical information; And can change at any time, need carry out real-time optimizing.
Transferring excellent variable is the set-point of the pressure of the drill in the pressure of the drill automatic control system.
Need to consider various constraints, fault and improper operating mode in the evolutionary process.Main constraint has the bound of the pressure of the drill, rig moment of torsion bound, the constraint of the pressure of the drill control system etc.Except that the fault of instrument and equipment, detect when occurring bouncing of drilling tool in the drilling process with other improper operating mode, in time stop to transfer excellent.Optimizing also can suspend by the optimizing logic of setting voluntarily.
Self-running judgement of optimizing and condition are: under the situation of the normal operation of the pressure of the drill automatic control system, if ROP, the rig moment of torsion; Mud flow rate and character have significant change, and fault or improper operating mode are removed, and have the data of following two the optimizing Cycle Lengths of normal operation conditions; Or the time-out optimizing of setting time is up; Or after the automatic control of the pressure of the drill automatic control system input, obtain in optimizing under the condition of production management slip-stick artist permission, optimizing is with self-starting.
Embodiment 2: the real-time optimization of catalytic cracking reaction process
Catalytic cracking reaction is that heavier cracking of oil is generated petroleum gas, liquid hydrocarbon, gasoline, the production run of low-density oil cuts such as diesel oil.Its optimization aim is the weighted sum of intermingled dregs ratio in above each reaction product productive rate and the raw material.
Because each yield of product can not measure in real time in the reaction product, the size of each productive rate changes because of the raw material oil composition is different, and the raw material oil composition also can not measure in real time.For this reason, adopt the real-time calculator of optimization aim provided by the invention, the measured value of each product flow that distillates based on the fractionating column that product is separated into each product and other actual measurement variablees; Based on dynamic mathematical models, consider that material and the heat amount of savings in fractionating column and container changes account temperature; The interior backflow; The oil gas dividing potential drop, the influence that (except that temperature, all can not survey, also need be calculated in real time by dynamic model) such as oil gas linear speeds separates product; Consider that reactor outlet to the time lag that distillates between the product measured discharge, finally calculates the productive rate of each product of reactor outlet under the standard separate condition.Simultaneously,, calculate the amount of burnt of regenerator, take into account time lag, provide the burnt rate of product of reactor based on the dynamic mathematical models and the actual measurement variable of regenerator.
What employing Chinese patent 200410048083.8 provided is the automatic control system of controlled variable with macroreaction heat, as transferring excellent controller.The controller based on model that adopts Chinese patent 99105546.2 or 200710118864.3 to provide is implemented the float area of macroreaction heat and temperature of reaction and is coordinated control, and maintenance reaction heat is optimal value, also keeps temperature of reaction in given bound.
The main constraint of evolutionary process has: the float area of macroreaction heat and temperature of reaction is coordinated the constraint of related variable in the control, the constraint of regenerator and main air blower, and the constraint of rich gas compressor, the constraint of fractionator, or the like.
The operating mode of self-starting optimizing detects and mainly contains: under this constant situation of macroreaction hot radical; The reaction product productive rate has significant change; Transfer excellent controller output that significant change is arranged; Fractionator and regenerator operation conditions have significant change, and the optimizing data of sufficient length are arranged after constraint or fault (fault that comprises the real-time counter of optimization aim) are removed.
Embodiment 3: the real-time optimization of catalytic reforming reaction process
Here be that master's catalytic reforming reaction is an example to produce aromatic hydrocarbons, its optimization aim is that aromatics yield is the highest.But aromatics yield can not measure in real time, and the optimization aim computing device among the present invention according to the mechanism and the characteristics of reforming reaction, utilizes dynamic mathematical models and actual measurement variable to provide solution to this problem, and following three approach is arranged:
(1) calculates with reaction heat and dynamic mathematical models: mainly contain the thermonegative reaction that generates aromatic hydrocarbons in the reforming reactor and be cracked into pentane or the more themopositive reaction of light constituent.The pentane that can separate by the reactor downstream depentanizer and the measured discharge of light constituent more; Utilize the dynamic mathematical models of fractionator and associated vessel to calculate the pentane of reactor outlet and the flow of light constituent more, and then obtain the cracking reaction liberated heat.Can calculate the grand tube reaction heat (showing as thermonegative reaction) of course of reaction by the dynamic mathematical models of reactor and actual measurement variable.Can calculate aromatics yield by macroreaction heat with the hot sum of cracking reaction.
(2) the hydrogen productive rate and the dynamic mathematical models that produce with reaction are calculated: the reaction that raw material generates aromatic hydrocarbons is the reaction that attaches hydrogen producing, and cracking reaction is the reaction of consuming hydrogen.Utilize downstream depentanizer dynamic mathematical models to calculate pentane productive rate and corresponding hydrogen-consuming volume, can calculate the amounts of hydrogen that the aromatic hydrocarbons reaction produces,, calculate aromatics yield according to stoichiometric coefficient by actual measurement hydrogen yield (adding dynamic correction).
(3) by the downstream extraction tower with aromatic hydrocarbons with after non-aromatic hydrocarbons separates, utilize dynabook to learn model and carry out dynamic compensation, calculate aromatics yield.
More than three kinds of methods all consider time lag based on actual conditions.
More than three kinds of result of calculations may not be very accurately, but as long as reacting phase is to changing the requirement that just can satisfy optimizing.The basis for estimation that available three's weighted sum changes as optimization aim.If wherein certain algorithm has fault or limits and can not calculate because of physical condition, can the weighting coefficient of this algorithm be made as zero.
What employing Chinese patent 200410048083.8 provided is the automatic control system of controlled variable with macroreaction heat, as transferring excellent controller.The controller based on model that adopts Chinese patent 99105546.2 or 200710118864.3 to provide is implemented the float area of macroreaction heat and temperature of reaction and is coordinated control, and maintenance reaction heat is optimal value, also keeps temperature of reaction in given bound.
Usually catalytic reforming reaction is made up of four tandem reactors, and the cracking reaction proportion is the highest in the 4th reactor.When calculating productive rate and optimization aim, calculate the macroreaction heat and the total reaction heat of each reactor by dynamic mathematical models.When the enforcement accent was excellent, intelligent optimizing device served as to transfer excellent variable with the hot set-point of first three reactor macroreaction.
Embodiment 4: the real-time optimization of pyrolysis furnace
Pyrolysis furnace is to be raw material with light or heavy petroleum fractions, is the process units of major product with ethene and propylene.Because the complicacy of petroleum fraction in the process of Pintsch process, also can generate hydrogen and other hydrocarbon products.Along with increasing of the cracking reaction degree of depth (mainly depending on and the cracking temperature and the residence time), the productive rate of ethene and propylene will descend after reaching mxm., and the productive rate of hydrogen and methane then constantly increases.The coke that cracking produces, make in the boiler tube with the quencher coking to the coke cleaning that to a certain degree need stop production, also be the key factor that influences benefit.The size of each product yield and relevant with feedstock property and operating conditions with the situation of change of cracking severity needs a real-time optimization device at any time pyrolysis furnace to be remained under the optimization situation and moves.
Below the optional usefulness of pyrolysis furnace optimization aim one of two kinds:
(1) optimization aim 1=C
2+ w
2C
3-w
sS
M
C
2=ethylene yield C
3=productivity of propylene w
3＞0, be weighting coefficient
M>=1 is setup parameter w
S＞0, be weighting coefficient ,-w
sS is the penalty function (2) to coking
Number of components in the N=pyrolysis product can be selected number of components as required
C
iThe productive rate w of i component in the=pyrolysis product
iThe weighting coefficient of=the i component
Pyrolysis gas on-line chromatograph analyser can provide methane, ethene and productivity of propylene data, and the productive rate in the above-mentioned optimization aim is according to chromatographic data, the pyrolysis furnace that calculates through dynamic mathematical models exports out each product productive rate.If comprise gasoline and other heavy product, pyrolysis furnace capable of using downstream production run measured data and dynamic mathematical models online in real time are calculated in the optimization aim 2.
The controller based on model that adopts Chinese patent 99105546.2 or 200710118864.3 to provide is the controlled variable of controller with cracking severity or pyrolysis furnace outlet temperature, forms to transfer excellent controller.With the set-point of cracking severity or pyrolysis furnace outlet temperature as transferring excellent variable.
Claims (1)
1. the real-time optimization device of a continuous flow procedure; It is characterized in that: by production run monitor (1), the real-time counter of optimization aim (2), intelligent optimizing device (3); Transfer excellent controller (4); Operation is kept watch on setting apparatus (5) five parts and is formed, and the real-time optimization device is obtained the service data of production run (7) in real time through existing data input/output interface (6), and will transfer excellent effect to give production run (7) and transfer excellent;
(1) production run monitor: have following function:
1). data processing and fault diagnosis: based on dynamic mathematical models result of calculation with the actual measurement variable poor, whether real-time judge operating condition and real measured data normal;
2). constraint forecast: according to dynamic model following variation of variable estimated, realized that variable reaches the forecast of bound, forecast stop when transfiniting the transfiniting accent of direction is excellent, to having estimated of variable following " simplification " and " complete " dual mode:
Simplified way: Y (t+p/t)=Y (t)+A [Y (t)-Y (t-1)] (1)
The t=current time, p=estimates time domain, the prediction coefficient that A=can be provided with
Y (t+p/t)=current time is to following t+p variable discreet value constantly
Y (t), the measured value of Y (t-1)=current time and previous moment variable
Complete mode: based on the discrete time state-space model,
Model: X (k+1)=F [X (k-τ
X), U (k-τ
U)] (2A)
Y(k)＝G[X(k)，U(k)] (2B)
Y=need estimate following variable X=state variable U=input variable
K=t/T
d=discrete time periodicity t=time T
d=discrete time the cycle
τ
XBetween=the state variable interactive retardation time corresponding discrete time periodicity
τ
U=input variable is to the forecast through the online in real time correction of the discrete time periodicity of the retardation time of state variable influence:
Y
C(k+p/k)＝Y(k+p/k)+[Y(k)-Y(k/k-p)] (3)
Y
C(k+p/k)=revised at following (k+p) T
dMoment predicted value
P is the corresponding discrete time periodicity of forecast time institute of setting
Y (k+p/k)=by (k+p) T in future of Model Calculation
dDiscreet value constantly
Y (k/k-p)=by the current time kT of Model Calculation
dDiscreet value
The measured value of Y (k)=current time k,
Y (k)-Y (k/k-p)=online in real time is to the modified value of model pre-estimating,
With revised predicted value Y
C(k+p/k) whether judgment variable transfinites;
3). working conditions change detects: normal in the production run operation, no variable is tied, and under the trouble-free situation, owing to can not survey the variation of variable, production run is moved when departing from the optimization situation, for the self-starting of optimizing provides information;
Working conditions change detects content and comprises: 1. relatively based on formula (2A) (2B) when each variable of calculating of state-space model and actual motion the size and the direction of each variable measured value whether significant change is arranged; 2. the heat and mass transport coefficient that calculates based on the dynamic mathematical models online in real time, concentration of component has significant change; 3. optimization target values has significant change; 4. operating condition has significant change, comprises that self-actuated controller or artificial performance variable have significant change, and production decision changes, and raw material is formed and fluctuations in discharge, the variable bound changed condition, and production equipment changes;
(2). the real-time counter of optimization aim: for the optimization aim that can not survey, according to production run mechanism, the actual measurement variable; Formula (2A) is the time lag of dynamic model and existence (2B); Online in real time is calculated and to be provided the optimization aim that can not survey, and when conversion ratio or the product in the reaction product that comprises chemical reactor was distributed as optimization aim, the measured discharge of each product of being separated by its downstream product separation equipment and the dynamic mathematical models of separation equipment were calculated the enter the mouth flow of each product of separation equipment; Thereby calculate corresponding optimization aim; During calculating, all to carry out Filtering Processing, to suppress noise to actual measurement variable and result of calculation;
When any the actual measurement variable that is used to calculate has fault or when improper, provides failure message, optimizing is suspended;
(3). intelligent optimizing device: realize the time-out and the self-starting of optimizing, the variable quantity of the excellent back of each accent desired value carried out dynamic compensation, satisfy stop to transfer when certain condition or operating mode do not allow excellent;
1) intelligent optimizing device " self-starting " optimizing: after not having constraint and possessing the optimizing data of enough time spans, the operator provides and transfers excellent order, or the operating mode judgement needs to transfer excellent; Or constraint release; Or the last time of setting transfer the excellent time out of accent after excellent to arrive, the optimizing device will start optimizing automatically, based on the change direction of transferring excellent variable and optimization aim; Confirm to transfer the adjustment direction and the size of excellent variable, transfer excellent; Transfer for the first time when excellent,, confirm the excellent direction of accent based on condition of setting and current working if transfer excellent variable or optimization aim no change; Each transfer excellent after, through the accent set after excellent cycle time, beginning for the second time and after each time accent excellent;
2) intelligent optimizing device transfers the variation of optimization target values in the excellent cycle to have dynamic compensation to each, prevents that wrong accent is excellent, shortens the excellent process of transferring; Dynamic compensation utilizes dynamic model, calculates when transferring excellent amount constant, transfers in the excellent cycle the 1st, 2 at one, up to the variation of N calculation time optimization aim, is called zero of optimization aim and responds; Use the optimization target values and zero respond poor of transferring excellent back practical manifestation to come out, as judging the foundation of transferring excellent direction;
3) in time stop to transfer excellent: multiple time-out standard is set; Make and transfer excellent process to stop; Comprising towards a direction transfers excellent number of times to reach setting value continuously; Twice excellent direction of accent is opposite; Transfer excellent back optimization aim not have significant change, transfer excellent suffering restraints, used real measured data or online in real time are calculated has fault; Abnormal condition appears in production process, and excellent data are transferred in the judgement that does not possess sufficient length;
(4). transfer excellent controller: the excellent effect of the accent that intelligent optimizer provides is implemented to transfer excellent through transferring excellent controller to production run;
(5). operation, keep watch on and setting apparatus: the operation of real-time optimization device is had only a run switch; Keep watch on and transfer excellent operation conditions, the function and the operational factor of real-time optimization device are carried out online adjustment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102895041A CN101957597B (en) | 2010-09-21 | 2010-09-21 | Real-time optimizer in continuous production process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102895041A CN101957597B (en) | 2010-09-21 | 2010-09-21 | Real-time optimizer in continuous production process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101957597A CN101957597A (en) | 2011-01-26 |
| CN101957597B true CN101957597B (en) | 2012-07-04 |
Family
ID=43484976
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102895041A Active CN101957597B (en) | 2010-09-21 | 2010-09-21 | Real-time optimizer in continuous production process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101957597B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102830616A (en) * | 2011-06-14 | 2012-12-19 | 北京三博中自科技有限公司 | Operation optimizing system and method of steam system |
| CN102393637B (en) * | 2011-10-14 | 2013-06-12 | 浙江大学 | Method for optimizing unit consumption of acetic acid of PTA (pure terephthalic acid) oxidation reactor in real time |
| CN103087751B (en) * | 2011-10-28 | 2016-01-20 | 中国石油化工股份有限公司 | A kind of construction method of industrial pyrolysis furnace value maximization model |
| CN103087753B (en) * | 2011-10-28 | 2014-12-03 | 中国石油化工股份有限公司 | Heavy raw material industrial cracking furnace value maximization model construction method |
| CN103087752B (en) * | 2011-10-28 | 2015-04-15 | 中国石油化工股份有限公司 | Industrial cracking furnace cracking product value maximization model construction method |
| CN103087755B (en) * | 2011-10-28 | 2015-04-15 | 中国石油化工股份有限公司 | Industrial cracking furnace cracking product value maximization model construction method |
| CN103087750B (en) * | 2011-10-28 | 2015-04-15 | 中国石油化工股份有限公司 | Naphtha industrial cracking furnace value maximization model construction method |
| CN105487493B (en) * | 2014-09-17 | 2017-12-01 | 中国石油化工股份有限公司 | Copper fusion process control method and system |
| CN112782979B (en) * | 2020-12-25 | 2022-05-03 | 杭州电子科技大学 | Real-time optimization control system and method for continuous catalytic reforming device |
| CN115386391B (en) * | 2022-08-01 | 2023-10-13 | 浙江中控技术股份有限公司 | Method and system for monitoring feeding flow of cracking furnace and processing faults |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1231436A (en) * | 1999-04-14 | 1999-10-13 | 袁璞 | Universal multi-variable quantity model pre-estimating coordinating control method |
| CN1713096A (en) * | 2004-06-15 | 2005-12-28 | 袁璞 | Control and control system for continuous chemical reactor production |
| CN101067743A (en) * | 2007-06-13 | 2007-11-07 | 袁璞 | Variable structural nonlinear model predictor controller |
-
2010
- 2010-09-21 CN CN2010102895041A patent/CN101957597B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1231436A (en) * | 1999-04-14 | 1999-10-13 | 袁璞 | Universal multi-variable quantity model pre-estimating coordinating control method |
| CN1713096A (en) * | 2004-06-15 | 2005-12-28 | 袁璞 | Control and control system for continuous chemical reactor production |
| CN101067743A (en) * | 2007-06-13 | 2007-11-07 | 袁璞 | Variable structural nonlinear model predictor controller |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101957597A (en) | 2011-01-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101957597B (en) | Real-time optimizer in continuous production process | |
| US20180122021A1 (en) | Chemical refinery performance optimization | |
| US11720072B2 (en) | Plant state operating analysis and control | |
| EP1904294B1 (en) | Catalyst withdrawal apparatus in a fluid catalyst cracking unit | |
| CN110187635B (en) | Method and apparatus for real-time optimization of continuous reformers | |
| CN102947423A (en) | Control of fluid catalytic cracking process for minimizing additive usage in the desulfurization of petroleum feedstocks | |
| CN103289725B (en) | Energy-saving optimization control method of ethylene cracking furnace | |
| CN103914595A (en) | Modeling method of medium-temperature coal tar total-fraction hydrogen cracking lumping kinetic model | |
| CN101414158A (en) | Method for optimizing cracking reaction operating condition of ethylene cracking furnace | |
| Pashikanti et al. | Predictive modeling of large-scale integrated refinery reaction and fractionation systems from plant data. Part 2: Fluid catalytic cracking (FCC) process | |
| CN105701267A (en) | Method for modelling oil catalytic cracking reaction regeneration part | |
| CN104463343A (en) | Method for predicting catalytic cracking light oil yield | |
| CN104392098A (en) | Method for predicting yield of catalytically cracked gasoline | |
| Han et al. | Modeling and optimization of a fluidized catalytic cracking process under full and partial combustion modes | |
| CN103524284A (en) | Forecasting and optimizing method for ethylene cracking material configuration | |
| CN1686973A (en) | Intelligent method for controlling cracking severity of cracking furnace in ethylene equipment | |
| CN101882239A (en) | Ethylene cracking severity modeling method based on expert knowledge and neutral network | |
| US20200108327A1 (en) | High Purity Distillation Process Control | |
| CN100340933C (en) | Control and control system for continuous chemical reactor production | |
| KR20080033419A (en) | Real-time operating optimized method of multi-input and multi-output continuous manufacture procedure | |
| Radu et al. | Modelling and simulation of an industrial fluid catalytic cracking unit | |
| RU2764214C1 (en) | Evaluation of the service life of the cyclone based on the remaining thickness of the abrasion-resistant lining | |
| US10953377B2 (en) | Delta temperature control of catalytic dehydrogenation process reactors | |
| CN101859103B (en) | On-line calculation and self-adaptation nonlinear prediction control method of catalytic cracking reaction depth | |
| Saleh et al. | A novel optimization formulation of fluid catalytic cracking unit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| PP01 | Preservation of patent right |
Effective date of registration: 20130830 Granted publication date: 20120704 |
|
| RINS | Preservation of patent right or utility model and its discharge | ||
| PD01 | Discharge of preservation of patent |
Date of cancellation: 20140723 Granted publication date: 20120704 |
|
| RINS | Preservation of patent right or utility model and its discharge | ||
| ASS | Succession or assignment of patent right |
Owner name: GENG XUESHAN Free format text: FORMER OWNER: YUAN PU Effective date: 20140909 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20140909 Address after: 100083 Beijing city Haidian District Wangzhuang Road No. 1, Tsinghua Tongfang Technology Plaza, B block 710 Patentee after: Geng Xueshan Address before: 100083 box 79, box 902, Beijing, Haidian Patentee before: Yuan Pu |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190419 Address after: 100083 Beijing Haidian District Wangzhuang Road No. 1 Courtyard No. 2 Building 6 Floor A Patentee after: Beijing Ruifei Huayi Technology Co., Ltd. Address before: 100083 Tsinghua Tongfang Science and Technology Plaza B, No. 1 Wangzhuang Road, Haidian District, Beijing, 710 Patentee before: Geng Xueshan |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20200810 Address after: 100083 Beijing city Haidian District Wangzhuang Road No. 1, Tsinghua Tongfang Technology Plaza, B block 710 Patentee after: Geng Xueshan Address before: 100083 Beijing Haidian District Wangzhuang Road No.1 Courtyard No.2 Building 6 Floor A Patentee before: Beijing Ruifei Huayi Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |