CN115454009A - Component distribution model predictive control method for chemical production - Google Patents
Component distribution model predictive control method for chemical production Download PDFInfo
- Publication number
- CN115454009A CN115454009A CN202211408678.4A CN202211408678A CN115454009A CN 115454009 A CN115454009 A CN 115454009A CN 202211408678 A CN202211408678 A CN 202211408678A CN 115454009 A CN115454009 A CN 115454009A
- Authority
- CN
- China
- Prior art keywords
- distillation range
- distillation
- temperature
- range point
- heating element
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000012824 chemical production Methods 0.000 title claims abstract description 11
- 238000009826 distribution Methods 0.000 title abstract description 13
- 238000004821 distillation Methods 0.000 claims abstract description 206
- 238000012549 training Methods 0.000 claims abstract description 64
- 238000006243 chemical reaction Methods 0.000 claims abstract description 56
- 238000004519 manufacturing process Methods 0.000 claims abstract description 24
- 230000015654 memory Effects 0.000 claims abstract description 9
- 238000010438 heat treatment Methods 0.000 claims description 58
- 238000009835 boiling Methods 0.000 claims description 56
- 239000000376 reactant Substances 0.000 claims description 15
- 230000003190 augmentative effect Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 14
- 230000003416 augmentation Effects 0.000 claims description 13
- 238000002360 preparation method Methods 0.000 claims description 10
- 230000004905 short-term response Effects 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 7
- 239000000203 mixture Substances 0.000 claims description 6
- 230000007787 long-term memory Effects 0.000 claims description 4
- 230000006403 short-term memory Effects 0.000 claims description 4
- 239000003153 chemical reaction reagent Substances 0.000 claims 1
- 230000004044 response Effects 0.000 claims 1
- 239000000126 substance Substances 0.000 abstract description 3
- 239000003054 catalyst Substances 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000005984 hydrogenation reaction Methods 0.000 description 5
- 150000004945 aromatic hydrocarbons Chemical class 0.000 description 3
- 239000002994 raw material Substances 0.000 description 3
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 2
- 239000003849 aromatic solvent Substances 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000001257 hydrogen Substances 0.000 description 2
- 229910052739 hydrogen Inorganic materials 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 230000000704 physical effect Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 238000010025 steaming Methods 0.000 description 2
- ZAMOUSCENKQFHK-UHFFFAOYSA-N Chlorine atom Chemical compound [Cl] ZAMOUSCENKQFHK-UHFFFAOYSA-N 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 150000001336 alkenes Chemical class 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003197 catalytic effect Effects 0.000 description 1
- 238000009903 catalytic hydrogenation reaction Methods 0.000 description 1
- 239000000460 chlorine Substances 0.000 description 1
- 229910052801 chlorine Inorganic materials 0.000 description 1
- 230000002301 combined effect Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013016 damping Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 229910001385 heavy metal Inorganic materials 0.000 description 1
- 231100000053 low toxicity Toxicity 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- JRZJOMJEPLMPRA-UHFFFAOYSA-N olefin Natural products CCCCCCCC=C JRZJOMJEPLMPRA-UHFFFAOYSA-N 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000035484 reaction time Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41865—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32252—Scheduling production, machining, job shop
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention relates to the technical field of chemical control, in particular to a component distribution model predictive control method for chemical production. The method obtains the reaction characteristics of each distillation range point, and matches the distillation range points in the distillation process under different control parameters according to the reaction characteristics to obtain a plurality of matched pairs. And further obtaining a switchable index according to the reaction characteristic difference between the distillation range points in the matching pair, selecting the matching pair based on the switchable index to construct an initial training set, and amplifying the initial training set according to the matching data to obtain a prediction model for training the long-short term memory production group to prepare, so as to predict real-time data and control container parameters according to the prediction parameters. The invention predicts the parameters to be adjusted of the future distillation range point by constructing the component distribution model, controls the parameters in time and ensures the production efficiency.
Description
Technical Field
The invention relates to the technical field of chemical control, in particular to a component distribution model predictive control method for chemical production.
Background
The high boiling point aromatic solvent is developed and produced by taking reformed aromatic hydrocarbon as a raw material according to international special aromatic solvent standards, and has the characteristics of strong dissolving power, low toxicity, small smell, high boiling point, slow volatilization, no water or olefin, no chlorine or heavy metal, stable chemical and physical properties, good leveling property and the like. Heavy aromatics are crude, different in properties in the former stage, and can extract 250 ℃ light fraction, and combine with tail oil with higher distillation range (about 260-380 ℃) to make reasonable use of C10 aromatics to produce high boiling point aromatics. In the general production process, catalytic hydrogenation is needed, and then rectification separation is carried out, but at present, the preparation parameters are adjusted by manual experience, the condition of unstable quality is easy to occur, and manual monitoring is always needed in the later period, so that a system for intelligently judging whether a process needs to be improved or not and carrying out hydrogenation intervention is needed in the component preparation process, so that the steaming efficiency is improved.
Disclosure of Invention
In order to solve the above technical problems, an object of the present invention is to provide a component distribution model predictive control method for chemical production, which adopts the following technical solutions:
the invention provides a component preparation model predictive control method for chemical production, which comprises the following steps:
collecting a set temperature value, a steam temperature and a heating element temperature in the container of each distillation range point in the distillation process; obtaining appearance characteristic mixed codes of reactants in the container; constructing a short-term reaction descriptor according to the steam temperature, the heating element temperature and the volume space velocity of the container in the sampling process;
matching different distillation range points according to appearance characteristic mixed coding difference and short-term reaction descriptor difference between the distillation range points in the distillation process under different control parameters to obtain a plurality of matched pairs;
obtaining a switchable index based on a set temperature value difference between two distillation range points in the matched pair, a steam temperature difference and a heating element temperature difference; selecting data in a preset number of matching pairs with the largest switchable indexes as an initial training set, amplifying the initial training set according to mutually matched distillation range data in the initial training set to obtain a training set, and training a long-term and short-term memory production component system prediction model according to the training set;
inputting the real-time steam temperature and the real-time heating element temperature of the real-time distillation range point in the real-time distillation process into the long-term and short-term memory production component preparation prediction model, obtaining the set temperature value and the volume airspeed corresponding to the next distillation range point, and controlling the container parameters according to the set temperature value and the volume airspeed corresponding to the next distillation range point.
Further, the obtaining of the appearance feature mix code of the reactant in the container comprises:
obtaining descriptive words of appearance characteristics of reactants in the container, and constructing TF-IDF appearance codes according to the descriptive words; collecting color information of the container reactant, and constructing an appearance color code according to component values in an RGB color space; and combining the TF-IDF appearance coding and the appearance color coding to obtain appearance characteristic mixed coding.
Further, the obtaining the appearance feature hybrid coding comprises:
expanding the dimension of the appearance color code to obtain a first expanded code; merging the first extended code and the TF-IDF appearance code to obtain a second extended code; and reducing the dimension of the second extended code to a preset dimension to obtain an appearance characteristic mixed code.
Further, the constructing a short-term response descriptor based on the steam temperature, the heating element temperature, and the volumetric space velocity of the vessel during sampling comprises:
the short-term reaction descriptor is composed of a steam temperature sequence median value, a heating element temperature sequence median value, a steam temperature sequence mean value, a heating element temperature sequence mean value, a steam temperature sequence initial value and a heating element temperature sequence initial value at each distillation point in the distillation process.
Further, the method for obtaining a plurality of matching pairs comprises:
constructing a matching distance according to appearance characteristic mixed coding difference and short-term reaction descriptor difference between different distillation processes under different reaction parameters; and matching different distillation processes by utilizing a KM matching algorithm according to the matching distance to obtain a plurality of matching pairs.
Further, the constructing the matching distance according to the appearance feature mixed coding difference and the short-term reaction descriptor difference between different distillation processes under different reaction parameters comprises:
obtaining a matching distance according to a matching distance formula, wherein the matching distance formula comprises:
wherein,
is the matching distance between boiling point X and boiling point Y,
is the short term reaction descriptor similarity between boiling point X and boiling point Y,
the similarity is codified for the appearance feature mixture between boiling point X and boiling point Y,
the absolute value of the difference of the distillation range point serial numbers corresponding to the distillation range point X and the distillation range point Y in the whole distillation process is shown; if it is
Then, then
Is infinite.
Further, the obtaining a switchable index based on the set temperature value difference, the steam temperature difference, and the heating element temperature difference in the two distillation processes in the matched pair comprises:
obtaining a switchable index according to a switchable index formula, the switchable index formula comprising:
wherein,
is a switchable index between the boiling point X and the boiling point Y,
is the absolute value of the difference in the set temperature value between boiling point X and boiling point Y,
is the matching distance between the boiling point X and the boiling point Y,
is the dynamic time-warping distance of the vapour temperature sequence between the boiling point X and the boiling point Y,
is the dynamic time-warping distance of the heating element temperature sequence between boiling point X and boiling point Y.
Further, the augmenting the initial training set according to distillation process data matched with each other in the initial training set comprises:
replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs to obtain first augmentation training data;
replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the next distillation range point of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs to obtain second augmentation training data;
respectively replacing the target distillation range point and the volume airspeed, set temperature, steam temperature and heating element temperature of the next distillation range point of the target distillation range point with the volume airspeed, set temperature, steam temperature and heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs, and obtaining third augmentation training data;
and combining the first augmented training data, the second augmented training data and the third augmented training data of each distillation range point to obtain a training set.
The invention has the following beneficial effects:
the embodiment of the invention is used for characterizing the reaction characteristics in the distillation process by constructing short-term reaction descriptors and appearance characteristic mixed codes. The distillation range points of the distillation process under different parameters can be matched according to the reaction characteristics, then the matching pairs with large switchable indexes are selected as training data, the data are expanded according to the data between the matching pairs, so that the subsequent training process of the long-short term memory production component distribution prediction model has enough training basis, the long-short term memory production component distribution prediction model can output an optimal result according to the expanded data, the referential property of the prediction data is ensured, the set temperature value and the volume airspeed at the future moment are predicted through the long-short term memory production component distribution prediction model, and the parameters in the container are controlled. The control on the container parameters in time is realized, and the production efficiency is ensured.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a group distribution model predictive control method for chemical production according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a color information collection process according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention for achieving the predetermined objects, the following detailed description of the components distribution model predictive control method for chemical production according to the present invention with reference to the accompanying drawings and preferred embodiments will be given below. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following describes a specific scheme of the component preparation model predictive control method for chemical production in detail with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a group distribution model predictive control method for chemical production according to an embodiment of the present invention is shown, where the method includes:
step S1: collecting a set temperature value, a steam temperature and a heating element temperature in the container of each distillation range point in the distillation process; obtaining appearance characteristic mixed codes of reactants in the container; the short-term reaction descriptors were constructed based on the steam temperature, heating element temperature and volumetric space velocity of the vessel at the sampling process.
According to the embodiment of the invention, CIO heavy aromatic hydrocarbon and CIO tail oil after cutting and extracting light components are mixed according to the volume ratio of 2. The component preparation production for performing hydrogenation and lightening is taken as an example, and the component preparation, namely the hydrogenation process, is distributed. In the embodiment of the invention, the hydrogen pressure is 5MPa, and for a certain catalytic process as an example, the temperature of steam distilled out is controlled in real time for the whole distillation process, and the regulated nodes are respectively 0% (initial distillation range), 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95% and final distillation point (dry point) of the total volume, namely, in the embodiment of the invention, 12 distillation range points exist in one distillation process. In the embodiment of the invention, the set temperature value is preset to 98 ℃ for the initial distillation range, and so on, the temperature generally gets higher and higher, and the limitation is not made here, that is, each distillation range point corresponds to one set temperature value.
The heating mode and the control mode of the distillation process under different reaction parameters are different, and the hydrogenation degree is different, so that the temperature information in the container under the distillation process needs to be measured in real time, and the temperature information in the container specifically comprises the steam temperature and the temperature of the heating element. Because there is a certain numerical hysteresis between the steam temperatures, that is, there is a combined effect of endothermic and specific heat capacity in the data between the temperature of the heating element of the reaction vessel and the steam temperature, and there is a certain hysteresis, that is, temperature damping, which is related to the characteristics of the heavy aromatic hydrocarbon feedstock, the rate of the hydrogenation reaction, and the content characteristics of the two-phase material, it is necessary to collect both the steam temperature and the heating element temperature when collecting the temperature information.
In an embodiment of the invention, the steam temperature T _ A and the heating element temperature T _ B within the reaction vessel are recorded at a rate of 1 Hz.
In the embodiment of the invention, a plurality of steam temperature record values are arranged between each distillation range point, and nearest neighbor resampling is carried out on the values to ensure that 400 data exist between each distillation range point. That is, the data of each distillation range point is processed into a fixed-length temperature record value, and the data is two fixed-length sequences of T _ A and T _ B. That is, each temperature sequence corresponding to each boiling point is a sequence consisting of 400 data point information.
During distillation, each distillation range point has different morphological characteristics of the reactants in the container due to the influence of reaction parameters, such as color, transparency and other information of the reactants. Therefore, in order to further describe the distillation process, it is also necessary to obtain an appearance characteristic mixed code of the reactants in the container at each distillation point in the distillation process, and the appearance characteristic mixed code is used to characterize the morphological information of the reactants at each distillation point, which specifically includes:
analyzing the physical property data, obtaining description words of the appearance characteristics of the reactants in the container, and expressing the description words through conventional words: the appearance words can be yellowish, clear and the like, in order to keep the words in the industry uniform and realize word disambiguation, in the embodiment of the invention, english words are used, for example, words are segmented in Bright & Clear to Bright and Clear, so that TF-IDF appearance codes after word segmentation based on a word stock are obtained. It should be noted that the dimension of the appearance coding is related to the size of the word stock of the segmented words, and the specific coding method is a technical means well known to those skilled in the art and is not described herein; the method for acquiring the descriptive words of the appearance characteristics of the reactants in the container can be acquired through visual observation or a neural network, is not limited herein, and the acquisition mode can be freely selected according to specific implementation scenes.
Further collecting color information of the container reactants, and constructing an appearance color code by component values in the RGB color space. Referring to fig. 2, a schematic diagram of a color information collection process according to an embodiment of the present invention is shown, in which an ADJD-S313-QR999RGB component chromaticity sensor is used to measure an external RGB value under a constant light source brightness, and a bypass is built in a reaction pipeline or an observation window is directly installed in the reaction pipeline. In the embodiment, an avago company CMOS digital color sensor with the model of add-s 313-qr999 is used, a light source 1 provides backlight to obtain a transmitted color reference, a light source 2 provides a diffuse reflection color reference, and after RGB chromaticity measurement, RGB components are normalized to three floating point values of 0 to 1 based on the maximum measurement range of the sensor, namely the appearance color code of each distillation range point is a code consisting of normalized data of three color channel components.
Further merging the TF-IDF appearance coding and the appearance color coding to obtain an appearance characteristic mixed coding, which specifically comprises the following steps:
a. firstly, performing dimension amplification on a plurality of normalized RGB components in the appearance color code C to obtain a 12 x 3 dimensional high dimension vector first extended code H _0.
b. And then the coded data is combined with the character TF-IDF appearance coding to obtain a second extended coding H _1 with the high dimensional vector of 12X 3+ X dimension, wherein X is the dimensional number of the TF-IDF appearance coding.
c. And further reducing the dimension of the second extended code to a preset dimension to obtain the appearance characteristic mixed code. In the embodiment of the invention, a high-dimensional space is constructed based on the high-dimensional vector H _1 in each production record, and the dimension is reduced to 10 dimensions based on a PCA algorithm to obtain appearance feature mixed codes. I.e. the preset dimension is 10 dimensions in the embodiment of the present invention.
i. Specifically, from the empirical value, 10 dimensions are the dimensional number with acceptable precision, the implementer can continue to increase the dimensions based on the dimensional number, so as to ensure the data precision, and the vector after dimension reduction is mainly used for representing the physical appearance data of each steaming process without control.
And ii, obtaining the appearance characteristic mixed code H after dimensionality reduction based on the previous data.
The appearance characteristic mixed code H has the function of jointly representing subjective information and objective information aiming at the distillation process, so that subjective errors of manual analysis are avoided, and meanwhile, the accuracy and the objectivity of data in the storage and management process are improved.
According to each production data, constructing a short-term reaction descriptor according to the steam temperature, the heating element temperature and the volume space velocity of the container in the sampling process, wherein the short-term reaction descriptor specifically comprises the following steps:
taking hydrogen pressure 6MPa as an example, the mixture ratio of the catalyst can be finely adjusted and the distillation temperature can be finely adjusted in the production process, correspondingly, taking the hydrogen-oil volume ratio 1000 as an example, the volume space velocity V can be between 1.8 and 1.5, and the distillation temperature at each distillation range point can have different target values. In general, a higher V indicates a higher catalyst activity that can be assumed, and a higher corresponding plant throughput. It should be noted that any reaction vessel and reaction principle have marginal effect, V can not be increased infinitely, for the measured vessel device, the larger V means more raw materials passing through the catalyst in unit time, the retention time of the raw materials on the catalyst is short, and the reaction depth is shallow; conversely, small V means long reaction times, and generally decreasing V is advantageous for increasing the conversion of the reaction. However, too low V means that a large amount of catalyst is required for the same treatment amount, and it is economically unreasonable to reduce the reaction efficiency by changing the phase in the distillation range. Therefore, V represents the experience after the fine adjustment of the catalyst at the time of production, and is not necessarily an optimum parameter. The temperature parameter and the volume space velocity V are therefore factors which influence the contextual characteristics of the short-term reaction. For V, the recorded V is generally tuned by fine tuning and experience since it is already in production. Therefore, the short-term reaction descriptor is constructed according to the steam temperature, the heating element temperature and the volume space velocity of the container under the sampling process, and specifically comprises the following steps:
the short-term reaction descriptor consists of a steam temperature sequence median value, a heating element temperature sequence median value, a steam temperature sequence mean value, a heating element temperature sequence mean value, a steam temperature sequence initial value and a heating element temperature sequence initial value at each distillation range point in the distillation process. That is, the short-term response descriptor is a vector composed of a plurality of characteristic parameters and is expressed in the form of mathematical expression
Wherein
Is the median value in the steam temperature sequence,
is the median value in the heating element temperature sequence,
is the average value of the steam temperature sequence,
is the mean value of the heating element temperature sequence,
is the initial value of the steam temperature sequence,
is the initial value of the heating element temperature sequence.
Step S2: and matching different distillation range points according to the appearance characteristic mixed coding difference and the short-term reaction descriptor difference between each distillation range point in the distillation process under different control parameters to obtain a plurality of matched pairs.
First, different volume space velocities V, catalyst adjustments and temperature adjustments will yield different postpartum data, so for the short-term reaction mode descriptor Q and the color reference C of the distillation range point, type assignments can be made based on a search of the K-M operator according to the following strategy: searching the similarity and similarity of each short-term reaction mode and the reaction mode of the next reaction distillation range, and constructing model parameters for the intervention degree of subsequent control and the temperature recommended value. Namely, the reaction process of each distillation range point is taken as a short-term reaction process, and the distillation range points in the distillation process under different reaction parameters are matched. For example, the 2 nd boiling point in the reaction mode a is set as a target boiling point, and the target boiling point is matched with all the boiling points in all other reaction modes such as the reaction mode B and the reaction mode D.
Obtaining a matching distance according to a matching distance formula, wherein the matching distance formula comprises:
wherein,
is the matching distance between the boiling point X and the boiling point Y,
short term reaction descriptor similarity between boiling point X and boiling point Y,
the similarity is coded for the appearance feature mixture between the boiling point X and the boiling point Y,
the absolute value of the difference of the corresponding distillation range point serial numbers of the distillation range point X and the distillation range point Y in the whole distillation process is shown; if it is
Then, then
Is infinite.
In a matching distance formula, multiplying the short-term reaction descriptor similarity and the appearance feature mixed coding similarity and then subtracting by a numerical value 1, namely
The integral difference of the two data is shown, and the difference of the distillation range point sequence numbers is further introduced, so that the distillation range points with different distillation range point sequence numbers can be matched in the subsequent matching process, the distillation range point sequence number difference of the two matched distillation range points is not too large, and the reference of subsequent predicted data is increased.
In the embodiment of the present invention, it is,
for cosine similarity between two short-term response descriptors,
cosine similarity between the two appearance features is mixed encoded. It should be noted that cosine similarity is well known in the prior art, and the cosine similarity between short-term response descriptors is taken as an example, and an expression of the cosine similarity is listed here:
wherein,
as short-term response descriptors
To (1)
The number of the elements is one,
as short-term response descriptors
To (1)
And (4) each element.
Under the distance constraint of the formula D, the K-M algorithm searches to obtain a plurality of matching pairs, and most of matching results among the matching pairs are that the distillation range nodes are different by 1, such as 10% to 20%; and in the short-term reaction mode, the temperature characteristics and the parameters of V are similar, and the appearance is similar.
And step S3: obtaining a switchable index based on a set temperature value difference between two boiling point values in the matched pair, a steam temperature difference, and a heating element temperature difference; selecting data in a preset number of matching pairs with the maximum switchable indexes as an initial training set, amplifying the initial training set according to mutually matched distillation range data in the initial training set to obtain a training set, and training a long-term and short-term memory production group according to the training set to prepare a prediction model.
The boiling point in one matching pair can represent the temperature value of the next boiling range that can be matched under different or as similar parameters as possible. Based on the temperature values and the similarity degree of the temperature information of each distillation range point, screening out a reference matching pair which is more suitable in the matching pair:
defining a switchable index for each matched pair, the index representing the extent of sustainable delayed distillation between the upper and lower boiling ranges, and if the index is larger, considering that the fraction producible in this case is more, i.e. the switchable index is obtained according to the set temperature value difference, the steam temperature difference and the heating element temperature difference in the two boiling ranges points in the matched pair, and specifically comprises:
the switchable index is obtained according to a switchable index formula, which includes:
wherein,
is a switchable index between boiling point X and boiling point Y,
is the absolute value of the difference between the set temperature values between the distillation range point X and the distillation range point Y,
is the matching distance between the boiling point X and the boiling point Y,
is the dynamic time-warping distance of the vapour temperature sequence between the boiling point X and the boiling point Y,
is the dynamic time-warping distance of the heating element temperature sequence between boiling point X and boiling point Y.
In the switchable index formula, the denominator is the sum of two distances, i.e. the larger the distance is, the larger the difference between two distillation range points is, the smaller the switchable index is; the product of the difference absolute value of the molecular position set temperature value and the inverse of the matching distance, namely the smaller the difference absolute value and the smaller the matching distance, shows that the closer the two distillation range points are, the larger the switchable index is.
Data in a preset number of matching pairs with the largest switchable index are selected as an initial training set, and in the embodiment of the invention, the matching pairs twenty-five percent of the top rank of the switchable index are selected to construct the initial training set.
For twenty-five percent of matching pairs before ranking, wherein any matching pair has a distillation range in sequence, and the initial training set is augmented according to postpartum data corresponding to the previous distillation range point, namely the initial training set is augmented according to distillation range data matched with each other in the initial training set to obtain the training set, and the specific augmentation method comprises the following steps:
replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs to obtain first augmentation training data; replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the next distillation range point of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs to obtain second augmentation training data; respectively replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the target distillation range point and the next distillation range point of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs, and obtaining third augmentation training data; and combining the first augmented training data, the second augmented training data and the third augmented training data of each distillation range point to obtain a training set.
For example, for the target distillation range point a, there is one distillation process A, and the distillation range point number in the distillation process A is i; the target distillation range point a corresponds to a matched distillation range point B, and the matched distillation range point B belongs to a distillation process B. Replacing the first augmentation training data of the target distillation range point a by distillation range point data with the distillation range point serial number i in the distillation process B; replacing the distillation range point data with the distillation range point serial number of i +1 in the distillation process A with the distillation range point data with the distillation range point serial number of i +1 in the distillation process B by the second augmentation training data; and the third augmentation training data is to replace the distillation range point data with the distillation range point sequence number i in the distillation process A with the distillation range point data with the distillation range point sequence number i in the distillation process B, and replace the distillation range point data with the distillation range point sequence number i +1 in the distillation process A with the distillation range point data with the distillation range point sequence number i +1 in the distillation process B.
Based on the matching effect of K-M, the parameters of the temperature characteristic and V in the short-term response mode are approximate, and the postpartum data with similar appearance are taken as data sources. By augmenting the fraction that can be produced in this case is more control model training data. The long-short term memory production component preparation prediction model can be based on error adjustment of a plurality of pieces of data which are expanded under the action of Batch Size, so that a better prediction result matched with the postpartum data is obtained. For the training set, the tag is set to T _ C and space velocity V for the next distillation range based on T _ A, T _ B for the previous distillation range.
And step S4: inputting the real-time steam temperature of the real-time distillation range point and the real-time heating element temperature in the real-time distillation process into the long-short term memory production component preparation prediction model to obtain the set temperature value and the volume airspeed in the next distillation process, and controlling the container parameters according to the set temperature value and the volume airspeed in the next distillation process.
Therefore, in the real-time production process, based on the real-time steam temperature and the real-time heating element temperature of a real-time distillation range point in the real-time distillation process, resampling is carried out, and then a long-short term memory production component allocation prediction model is input, so that the set temperature value and the volume airspeed corresponding to the next distillation range point can be obtained, and the predictive optimization control effect can be achieved through manual setting of workers or automatic setting of a distillation container, so that the distillation range fraction output quantity is improved.
In summary, the embodiments of the present invention obtain the reaction characteristics of each distillation range point, and match the distillation range points in the distillation process under different control parameters according to the reaction characteristics to obtain a plurality of matching pairs. And further acquiring switchable indexes according to the difference of reaction characteristics between the distillation range points in the matching pairs, selecting the matching pairs based on the switchable indexes to construct an initial training set, and amplifying the initial training set according to matching data to acquire a prediction model for training the long-short term memory production components by the training set, so as to predict real-time data and control container parameters according to the prediction parameters. According to the embodiment of the invention, the parameters required to be adjusted for predicting the future distillation range point are constructed by constructing the component distribution model, so that the production efficiency is ensured.
It should be noted that: the sequence of the above embodiments of the present invention is only for description, and does not represent the advantages or disadvantages of the embodiments. The processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and should not be taken as limiting the scope of the present invention, which is intended to cover any modifications, equivalents, improvements, etc. within the spirit and scope of the present invention.
Claims (8)
1. A method for predictive control of a component dispensing model for chemical production, the method comprising:
collecting the set temperature value, the steam temperature and the temperature of a heating element in a container of each distillation range point in the distillation process; obtaining appearance characteristic mixed codes of reactants in the container; constructing a short-term reaction descriptor according to the steam temperature, the heating element temperature and the volume space velocity of the container in the sampling process;
matching different distillation range points according to appearance characteristic mixed coding difference and short-term reaction descriptor difference between the distillation range points in the distillation process under different control parameters to obtain a plurality of matched pairs;
obtaining a switchable index based on a set temperature value difference between two boiling point values in the matched pair, a steam temperature difference, and a heating element temperature difference; selecting data in a preset number of matching pairs with the largest switchable indexes as an initial training set, amplifying the initial training set according to mutually matched distillation range data in the initial training set to obtain a training set, and training a long-term and short-term memory production component system prediction model according to the training set;
inputting the real-time steam temperature of the real-time distillation range point and the real-time heating element temperature in the real-time distillation process into the long-short term memory production component preparation prediction model to obtain the set temperature value and the volume airspeed corresponding to the next distillation range point, and controlling the container parameters according to the set temperature value and the volume airspeed corresponding to the next distillation range point.
2. The method as claimed in claim 1, wherein the obtaining of the appearance feature mixture codes of the reagents in the containers comprises:
obtaining descriptive words of appearance characteristics of reactants in the container, and constructing TF-IDF appearance codes according to the descriptive words; collecting color information of the container reactant, and constructing an appearance color code according to component values in an RGB color space; and combining the TF-IDF appearance coding and the appearance color coding to obtain the appearance characteristic mixed coding.
3. The method as claimed in claim 2, wherein the obtaining of the appearance feature hybrid coding comprises:
expanding the dimension of the appearance color code to obtain a first expanded code; merging the first extended code and the TF-IDF appearance code to obtain a second extended code; and reducing the dimension of the second extended code to a preset dimension to obtain an appearance characteristic mixed code.
4. The method as claimed in claim 1, wherein the step of constructing the short-term response descriptor according to the steam temperature, the heating element temperature and the volume space velocity of the container during sampling comprises:
the short-term reaction descriptor is composed of a steam temperature sequence median value, a heating element temperature sequence median value, a steam temperature sequence mean value, a heating element temperature sequence mean value, a steam temperature sequence initial value and a heating element temperature sequence initial value at each distillation range point in the distillation process.
5. The method of claim 1, wherein the step of obtaining a plurality of matching pairs comprises:
constructing a matching distance according to appearance characteristic mixed coding difference and short-term reaction descriptor difference between different distillation processes under different reaction parameters; and matching different distillation processes by utilizing a KM matching algorithm according to the matching distance to obtain a plurality of matching pairs.
6. The method as claimed in claim 5, wherein the step of constructing the matching distance according to the appearance feature mixed coding difference and short-term response descriptor difference between different distillation processes under different response parameters comprises:
obtaining a matching distance according to a matching distance formula, wherein the matching distance formula comprises:
wherein,
is the matching distance between the boiling point X and the boiling point Y,
short term reaction descriptor similarity between boiling point X and boiling point Y,
the similarity is coded for the appearance feature mixture between the boiling point X and the boiling point Y,
the absolute value of the difference of the corresponding distillation range point serial numbers of the distillation range point X and the distillation range point Y in the whole distillation process is shown; if it is
Then, then
Is infinite.
7. The method of claim 1, wherein the obtaining the switchable index according to the set temperature value difference, the steam temperature difference and the heating element temperature difference in the two distillation processes in the matching pair comprises:
obtaining a switchable index according to a switchable index formula, the switchable index formula comprising:
wherein,
is a switchable index between the boiling point X and the boiling point Y,
is the absolute value of the difference between the set temperature values between the distillation range point X and the distillation range point Y,
is the matching distance between boiling point X and boiling point Y,
is the dynamic time-warping distance of the steam temperature sequence between the distillation range point X and the distillation range point Y,
is the dynamic time-warping distance of the heating element temperature sequence between boiling point X and boiling point Y.
8. The method of claim 1, wherein the augmenting the initial training set based on distillation process data that match each other in the initial training set comprises:
replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs to obtain first augmentation training data;
replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the next distillation range point of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs to obtain second augmentation training data;
respectively replacing the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the target distillation range point and the next distillation range point of the target distillation range point with the volume airspeed, the set temperature, the steam temperature and the heating element temperature of the corresponding distillation range point in the distillation process to which the matched distillation range point belongs, and obtaining third augmentation training data;
and combining the first augmented training data, the second augmented training data and the third augmented training data of each distillation range point to obtain a training set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211408678.4A CN115454009A (en) | 2022-11-11 | 2022-11-11 | Component distribution model predictive control method for chemical production |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211408678.4A CN115454009A (en) | 2022-11-11 | 2022-11-11 | Component distribution model predictive control method for chemical production |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115454009A true CN115454009A (en) | 2022-12-09 |
Family
ID=84295606
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211408678.4A Pending CN115454009A (en) | 2022-11-11 | 2022-11-11 | Component distribution model predictive control method for chemical production |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115454009A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102393636A (en) * | 2011-09-22 | 2012-03-28 | 任季明 | Control system and method for petroleum refining process |
| CN104152179A (en) * | 2014-09-01 | 2014-11-19 | 江苏联东化工股份有限公司 | Production method of high-boiling-point aromatic solvent oil |
| CN105844667A (en) * | 2016-03-25 | 2016-08-10 | 中国矿业大学 | Structural target tracking method of compact color coding |
| CN109726845A (en) * | 2017-10-31 | 2019-05-07 | 中国石油化工股份有限公司 | Product yield prediction technique, method for establishing model and the storage equipment being hydrocracked |
| CN109726844A (en) * | 2017-10-31 | 2019-05-07 | 中国石油化工股份有限公司 | Product yield method for automatically regulating, system and the storage equipment of hydrocracking unit |
| CN109754113A (en) * | 2018-11-29 | 2019-05-14 | 南京邮电大学 | Load forecasting method based on dynamic time warping Yu length time memory |
| CN110102075A (en) * | 2019-03-18 | 2019-08-09 | 山东金特昂莱测控技术有限公司 | The chlorination control method and device that view-based access control model identifies in bromine distillation technique |
| CN113764046A (en) * | 2021-09-24 | 2021-12-07 | 中国石油化工股份有限公司 | Method for improving yield of high value-added product in catalytic diesel oil hydroconversion |
-
2022
- 2022-11-11 CN CN202211408678.4A patent/CN115454009A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102393636A (en) * | 2011-09-22 | 2012-03-28 | 任季明 | Control system and method for petroleum refining process |
| CN104152179A (en) * | 2014-09-01 | 2014-11-19 | 江苏联东化工股份有限公司 | Production method of high-boiling-point aromatic solvent oil |
| CN105844667A (en) * | 2016-03-25 | 2016-08-10 | 中国矿业大学 | Structural target tracking method of compact color coding |
| CN109726845A (en) * | 2017-10-31 | 2019-05-07 | 中国石油化工股份有限公司 | Product yield prediction technique, method for establishing model and the storage equipment being hydrocracked |
| CN109726844A (en) * | 2017-10-31 | 2019-05-07 | 中国石油化工股份有限公司 | Product yield method for automatically regulating, system and the storage equipment of hydrocracking unit |
| CN109754113A (en) * | 2018-11-29 | 2019-05-14 | 南京邮电大学 | Load forecasting method based on dynamic time warping Yu length time memory |
| CN110102075A (en) * | 2019-03-18 | 2019-08-09 | 山东金特昂莱测控技术有限公司 | The chlorination control method and device that view-based access control model identifies in bromine distillation technique |
| CN113764046A (en) * | 2021-09-24 | 2021-12-07 | 中国石油化工股份有限公司 | Method for improving yield of high value-added product in catalytic diesel oil hydroconversion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107451101B (en) | Method for predicting concentration of butane at bottom of debutanizer by hierarchical integrated Gaussian process regression soft measurement modeling | |
| CN111738482A (en) | Method for adjusting technological parameters in polyester fiber polymerization process | |
| CN108760668A (en) | Pinus massoniana Seedlings root moisture method for fast measuring based on weighting autocoder | |
| CN114118622A (en) | Data trend prediction method and system based on time series | |
| Shokri et al. | Improvement of the prediction performance of a soft sensor model based on support vector regression for production of ultra-low sulfur diesel | |
| Hmamouche et al. | A causality based feature selection approach for multivariate time series forecasting | |
| Sheng et al. | Soft sensor design based on phase partition ensemble of LSSVR models for nonlinear batch processes | |
| He et al. | Weighted incremental minimax probability machine-based method for quality prediction in gasoline blending process | |
| Caetano et al. | Geographical classification of olive oils by the application of CART and SVM to their FT‐IR | |
| CN115048539A (en) | Social media data online retrieval method and system based on dynamic memory | |
| CN117453897B (en) | Document question-answering method and system based on large model and genetic algorithm | |
| CN115454009A (en) | Component distribution model predictive control method for chemical production | |
| Hartanto et al. | Stock Price Time Series Data Forecasting Using the Light Gradient Boosting Machine (LightGBM) Model | |
| CN115239613A (en) | Full-field digital slice image classification modeling method and device based on integrated deep learning | |
| Peng et al. | Multi-modal hybrid modeling strategy based on Gaussian mixture variational autoencoder and spatial–temporal attention: application to industrial process prediction | |
| CN117037931A (en) | Leaf chlorophyll content calculation method based on double-difference ratio index | |
| Emami et al. | Condensed-gradient boosting | |
| CN113723015A (en) | Catalytic cracking unit optimization method based on mechanism model and big data technology | |
| Krier et al. | Supervised variable clustering for classification of NIR spectra. | |
| CN110308240A (en) | A kind of electronic nose method for quickly identifying | |
| CN117973639B (en) | Photovoltaic power prediction method based on multiscale similarity days and improved integration algorithm | |
| JP7577934B2 (en) | Physical property prediction device, physical property prediction method, and manufacturing method | |
| CN112925202B (en) | Fermentation process stage division method based on dynamic feature extraction | |
| CN118626635B (en) | Sub-graph representation learning-based scientific research result clustering recommendation method and system | |
| CN114330562B (en) | Small sample refinement classification and multi-classification model construction method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |