CN115032891A - Polycrystalline silicon reduction furnace control method based on time series prediction - Google Patents
Polycrystalline silicon reduction furnace control method based on time series prediction Download PDFInfo
- Publication number
- CN115032891A CN115032891A CN202210958110.3A CN202210958110A CN115032891A CN 115032891 A CN115032891 A CN 115032891A CN 202210958110 A CN202210958110 A CN 202210958110A CN 115032891 A CN115032891 A CN 115032891A
- Authority
- CN
- China
- Prior art keywords
- time
- period
- reduction furnace
- current
- cooling water
- 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.)
- Granted
Links
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
- G05B11/00—Automatic controllers
- G05B11/01—Automatic controllers electric
- G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
- G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P.I., P.I.D.
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention relates to the field of control of polysilicon reducing furnace holes, and discloses a polysilicon reducing furnace control method based on time series prediction, wherein the resistance of a reducing furnace for a period of time in the future is predicted through an LSTM neural network, and when the predicted result shows that the resistance of the reducing furnace is about to deviate from an optimal resistance curve, the current of a silicon rod is controlled through a fuzzy PID controller without depending on a temperature sensor in the reducing furnace; the future state in the reduction furnace can be predicted in advance, so that the regulation and control hysteresis is reduced; regulation and control is based on sensor data, and the precision is higher than manual observation.
Description
Technical Field
The invention relates to the field of control of polysilicon reduction furnace holes, in particular to a polysilicon reduction furnace control method based on time series prediction.
Background
Polycrystalline silicon, as one of the important raw materials for the production of photovoltaic devices, is a cornerstone of the photovoltaic industry. In recent years, with the development of new energy industry, the production of polysilicon is more and more emphasized. The current mainstream method for producing the polycrystalline silicon is an improved Siemens method, which utilizes the principle of chemical vapor deposition, and when the temperature of a reduction furnace is about 1100 ℃, hydrogen and trichlorosilane generate vapor deposition reaction on a pretreated silicon rod to generate the polycrystalline silicon. The temperature of the reduction furnace is maintained by electrifying the silicon rod to generate heat. During the reaction, as the produced silicon is deposited on the surface of the silicon rod, the resistance of the silicon rod is changed, and the current of the silicon rod is adjusted correspondingly in order to keep the temperature in the reducing furnace stable. The quality of current control directly influences the power consumption of the production of the reduction furnace and the quality of products. For example, if the current value is large, the temperature in the furnace is high, which not only increases the energy consumption, but also degrades the quality of the polysilicon produced. Due to equipment, in an actual production scene, the temperature in the reduction furnace cannot be measured, the current is adjusted mainly by observing the silicon rod condition through an observation window manually, or the current resistance deviates from the actual optimal resistance and then is adjusted, the accuracy of the current resistance is poor, the current resistance has certain hysteresis, and the quality of a product is influenced when the resistance deviates from an optimal resistance curve. At present, some control methods based on actual states often need the temperature in the furnace as a feedback variable and cannot adapt to a reduction furnace without a temperature sensor.
Disclosure of Invention
In order to solve the technical problem, the invention provides a polycrystalline silicon reduction furnace control method based on time series prediction.
In order to solve the technical problems, the invention adopts the following technical scheme:
a polycrystalline silicon reduction furnace control method based on time series prediction comprises the following steps:
the method comprises the following steps: in each production period during normal operation of the polycrystalline silicon reduction furnace, every other first period T 1 Collecting primary production data; production data includes hydrogen flow rate v H2 Trichlorosilane flow v TCS Running time, reduction furnace tail gas temperature, cooling water rising temperature difference delta T and cooling water flow v water And the voltage U and the current I at the two ends of the silicon rod;
step two: calculating to obtain a second period T 2 Time series C of total internal hydrogen consumption H2 And a second period T 2 Time series C of total consumption of internal trichlorosilane TCS ;,n 1 Is a set integer;
step three: according to the cooling water temperature rise delta T and the cooling water flow v water Obtaining a second period T 2 A time sequence E of energy absorption of the internal cooling water;
step four: calculating the resistance R = U/I of the silicon rod according to the voltage U and the current I at the two ends of the silicon rod; converting the reduction furnace tail gas temperature and the silicon rod resistance R from the time sequence of the first period level to the time sequence of the second period level;
step five: taking the time sequence from the second step to the fourth step as training data to train the LSTM neural network, wherein the input of the training is the current time and T before the current time 3 The training data in time is output as T after the current time 4 Silicon rod resistance over time;,,is a set integer;
step six: recording the production period in which the product quality reaches the standard and the energy consumption is at the lowest first 30% in all production periods of the polycrystalline silicon reduction furnace as a production period B, and averaging the silicon rod resistors R at the same moment in each period in the production period B to obtain an optimal resistance curve;
step seven: averaging the silicon rod current I at the same time in each period in the production period B to obtain an optimal current curve, taking the optimal current curve as a reference curve controlled by the polycrystalline silicon reduction furnace, and taking the current at the corresponding time on the optimal current curve as the reference current I in the production process of the reduction furnace 1 ;
Step eight: calculating T output in the step five 4 Average value R of the resistance of the silicon rod over time 1 And calculating the average resistance R of the optimal resistance curve corresponding to the time 2 Calculating a difference R between the two 1 -R 2 ;
Step nine: will be different from R 1 -R 2 As input of the fuzzy PID controller, the corrected current I is obtained by the fuzzy PID controller 2 ;
Step ten: in the process of operating the polycrystalline silicon reduction furnace, every fifth period T 5 Repeating the seven to nine steps once to obtain the corrected current I 2 Will correct the current I 2 And a reference current I 1 Adding as final control current;,n 4 is a set integer.
Specifically, if one or several production data in step one is missing, the missing production data is replaced by an average of adjacent production data.
In particular, a first period T 1 =1s;n 1 =60, i.e. the second period T 2 =1min;n 2 =n 3 =n 4 =30, i.e. T 3 =T 4 =T 5 =30min。
Specifically, the second period T 2 =1 min; calculating the time sequence C of the total hydrogen consumption per minute in the second step H2 Then, the volume of hydrogen consumed in the ith minute is obtained by integrating the hydrogen flow rate by time:
WhereinThe flow rate of hydrogen in ith minute and jth second is(ii) a Adding the volumes of hydrogen input in every minute in the first k minutes to obtain the total hydrogen consumption in the first k minutes:
Taking different k values to time seriesAnd n is the time consumed by the final reaction of the polysilicon reduction furnace.
Specifically, the second period T 2 =1 min; calculating the time sequence C of the total trichlorosilane consumption in each minute in the second step TCS When the flow of the trichlorosilane is multiplied by the time to obtain the volume of the trichlorosilane consumed in the ith minute:
WhereinThe flow rate of trichlorosilane in ith minute and jth second is shown in unit of(ii) a Inputting the first k minutes per minute of the trichlorosilaneAdding the products to obtain the total consumption of the trichlorosilane in the previous k minutes:
Taking different k values to time seriesAnd n is the time consumed by the final reaction of the polysilicon reduction furnace.
Specifically, the second period T 2 =1 min; in the third step, the temperature difference Delta T is increased according to the cooling water and the flow v of the cooling water water And when the time sequence E of the energy absorbed by the cooling water in each minute is obtained, the energy E absorbed by the cooling water in the ith minute is calculated i :
Wherein c is the specific heat capacity of cooling water, rho is the density of the cooling water, and V water The cooling water flow at the ith minute, and the delta T are the temperature rise difference of the cooling water at the ith minute, and finally, the time sequence is obtained。
Compared with the prior art, the invention has the beneficial technical effects that:
the resistance of the reduction furnace for a period of time in the future is predicted through the LSTM neural network, and when the predicted result shows that the resistance of the reduction furnace is about to deviate from an optimal resistance curve, the current of the silicon rod is controlled through a fuzzy PID controller; the control method of the invention does not depend on a temperature sensor in the reduction furnace, and can predict the future state in the reduction furnace in advance, thereby reducing the lag of regulation and control; regulation and control is based on sensor data, and the precision is higher than manual observation.
Drawings
FIG. 1 is an overall flowchart of the control method of the present invention.
Detailed Description
A preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings.
A polycrystalline silicon reduction furnace control method based on time series prediction comprises the following steps:
s1: acquiring production data per second in each production cycle when the polycrystalline silicon reduction furnace normally operates, and using the production data per second for training an LSTM neural network; production data includes hydrogen flow rate v H2 Trichlorosilane flow v TCS Running time, reduction furnace tail gas temperature, cooling water rising temperature difference delta T and cooling water flow v water And the voltage U and the current I at the two ends of the silicon rod.
If a few seconds of production data are missing due to equipment anomalies, communication anomalies, or other causes, the average of the nearby production data can be substituted.
S2: calculating to obtain a time series C of the total hydrogen consumption per minute H2 And time series C of total trichlorosilane consumption per minute TCS 。
To calculate C H2 For example, the volume of hydrogen consumed at the i minute is obtained by integrating the hydrogen flow rate by time:
WhereinThe flow rate of hydrogen in ith minute and jth second is(ii) a Adding the volumes of hydrogen input in every minute in the first k minutes to obtain the total hydrogen consumption in the first k minutes:
Taking different k values to time seriesWherein n is the time consumed by the final reaction of the polycrystalline silicon reduction furnace;
the time sequence C of the total consumption amount of trichlorosilane per minute is obtained by the same method TCS :
Integrating the flow of trichlorosilane by the time to obtain the volume of trichlorosilane consumed in the ith minute:
WhereinIs the flow rate of trichlorosilane in ith minute and jth second and the unit is(ii) a Adding the volume of trichlorosilane input every minute in the first k minutes to obtain the total trichlorosilane consumption in the first k minutes:
Taking different k values to time seriesAnd n is the time consumed by the final reaction of the polysilicon reduction furnace.
S3: according to the cooling water temperature rise delta T and the cooling water flow v water Obtaining a second period T 2 The time sequence E for absorbing energy by the internal cooling water comprises the following specific steps: measuring the average value of the temperature difference of the cooling water in the same minute and the flow of the cooling water; calculating the energy E absorbed by the cooling water at the ith minute i :
Wherein c is the specific heat capacity of cooling water, rho is the density of the cooling water, and V water The volume of cooling water consumed for the ith minute is the cooling water flow, and Δ T is the temperature rise difference of the cooling water for the ith minute, and finally the time sequence is obtained。
S4: calculating the resistance R = U/I of the silicon rod according to the voltage U and the current I at the two ends of the silicon rod; and changing the two second-level time series of the silicon rod resistance and the reduction furnace tail gas temperature into a minute-level time series by averaging the production data in the same minute.
S5: training the LSTM neural network by taking the time sequence from S2 to S4 as training data, wherein the inputs of the training are the total hydrogen consumption per minute at the current moment and in the previous 30 minutes, the total trichlorosilane consumption per minute, the tail gas temperature of the reduction furnace, the running time, the absorption energy of cooling water per minute and the silicon rod resistance; the output is the resistance of the silicon rod for the next 30 minutes.
S6: and averaging the silicon rod resistors R at the same time in each period in all the production periods of the polycrystalline silicon reduction furnace to obtain an optimal resistance curve, wherein the production periods of the polycrystalline silicon reduction furnace in which the product quality reaches the standard and the energy consumption is lower by the first 30%.
S7: averaging the silicon rod currents I screened out in the S6 production cycle at the same moment in each cycle to obtain an optimal current curve, and taking the optimal current curve as polycrystalline silicon reductionReference curve of furnace control, in the production process of reduction furnace, the current corresponding to time on the optimum current curve is used as reference current I 1 ;
S8: and comparing the average value of the silicon rod resistance obtained in the step S5 at the future 30 minutes with the average resistance of the optimal resistance curve in the step S6 at the corresponding time, and calculating the difference between the two values.
S9: the difference obtained in S8 is used as the input of a fuzzy PID controller, and the corrected current I is obtained through the fuzzy PID controller 2 。
Fuzzy PID control is an existing control theory combining a PID algorithm and a fuzzy control theory. The fuzzy PID controller consists of two parts, namely a traditional PID controller and a fuzzification module. The parameters of the fuzzy PID controller are set according to the actual field condition.
S10: repeating the steps from S7 to S9 every 30 minutes in the operation process of the polycrystalline silicon reduction furnace to obtain the corrected current I 2 Will correct the current I 2 And a reference current I 1 And added as the final control current.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein, and any reference signs in the claims are not intended to be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (6)
1. A polycrystalline silicon reduction furnace control method based on time series prediction comprises the following steps:
the method comprises the following steps: in each production period during normal operation of the polycrystalline silicon reduction furnace, every other first period T 1 Collecting primary production data; production data includes hydrogen flow rate v H2 Trichlorosilane flow v TCS Running time, reduction furnace tail gas temperature, cooling water rising temperature difference delta T and cooling water flow v water And the voltage U and the current I at the two ends of the silicon rod;
step two: calculating to obtain a second period T 2 Time series C of total amount of internal hydrogen consumption H2 And a second period T 2 Time series C of total consumption of internal trichlorosilane TCS ;,n 1 Is a set integer;
step three: according to the cooling water temperature rise delta T and the cooling water flow v water Obtaining a second period T 2 A time sequence E of energy absorption of the internal cooling water;
step four: calculating the resistance R = U/I of the silicon rod according to the voltage U and the current I at the two ends of the silicon rod; converting the reduction furnace tail gas temperature and the silicon rod resistance R from the time sequence of the first period level to the time sequence of the second period level;
step five: taking the time sequence from the second step to the fourth step as training data to train the LSTM neural network, wherein the input of the training is the current time and T before the current time 3 The training data in time is output as T after the current time 4 Silicon rod resistance over time;,,is a set integer;
step six: recording the production period in which the product quality reaches the standard and the energy consumption is at the lowest first 30% in all production periods of the polycrystalline silicon reduction furnace as a production period B, and averaging the silicon rod resistors R at the same moment in each period in the production period B to obtain an optimal resistance curve;
step seven: averaging the silicon rod current I at the same time in each period in the production period B to obtain an optimal current curve, taking the optimal current curve as a reference curve controlled by the polycrystalline silicon reduction furnace, and taking the current at the corresponding time on the optimal current curve as the reference current I in the production process of the reduction furnace 1 ;
Step eight: calculating T output in the step five 4 Average value R of the resistance of the silicon rod over time 1 And calculating the average resistance R of the optimal resistance curve corresponding to the time 2 Calculating a difference R between the two 1 -R 2 ;
Step nine: will be different from R 1 -R 2 As the input of the fuzzy PID controller, the corrected current I is obtained by the fuzzy PID controller 2 ;
2. The polycrystalline silicon reduction furnace control method based on time series prediction according to claim 1, characterized in that: if one or several production data in step one are missing, the missing production data is replaced by an average of adjacent production data.
3. The method of claim 1The polycrystalline silicon reduction furnace control method based on time series prediction is characterized by comprising the following steps: first period T 1 =1s;n 1 =60, i.e. second period T 2 =1min;n 2 =n 3 =n 4 =30, i.e. T 3 =T 4 =T 5 =30min。
4. The polycrystalline silicon reduction furnace control method based on time series prediction according to claim 1, characterized in that: second period T 2 =1 min; calculating the time sequence C of the total hydrogen consumption per minute in the second step H2 Then, the volume of hydrogen consumed in the ith minute is obtained by integrating the hydrogen flow rate by the time:
WhereinThe flow rate of hydrogen in ith minute and jth second is(ii) a Adding the volumes of hydrogen input in every minute in the first k minutes to obtain the total hydrogen consumption in the first k minutes:
5. The polycrystalline silicon reduction furnace control method based on time series prediction according to claim 1 or 4, characterized in that: second period T 2 =1 min; calculating the time sequence C of the total trichlorosilane consumption in each minute in the second step TCS When the flow of the trichlorosilane is multiplied by the time to obtain the volume of the trichlorosilane consumed in the ith minute:
WhereinIs the flow rate of trichlorosilane in ith minute and jth second and the unit is(ii) a Adding the volumes of the trichlorosilane input every minute in the first k minutes to obtain the total consumption of the trichlorosilane in the first k minutes:
6. The time-based sequencing of claim 1The column prediction control method for the polycrystalline silicon reduction furnace is characterized by comprising the following steps: second period T 2 =1 min; in the third step, the temperature difference Delta T is increased according to the cooling water and the flow v of the cooling water water And when the time sequence E of the energy absorbed by the cooling water in each minute is obtained, the energy E absorbed by the cooling water in the ith minute is calculated i :
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210958110.3A CN115032891B (en) | 2022-08-11 | 2022-08-11 | Polycrystalline silicon reduction furnace control method based on time series prediction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210958110.3A CN115032891B (en) | 2022-08-11 | 2022-08-11 | Polycrystalline silicon reduction furnace control method based on time series prediction |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115032891A true CN115032891A (en) | 2022-09-09 |
CN115032891B CN115032891B (en) | 2022-11-08 |
Family
ID=83130112
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210958110.3A Active CN115032891B (en) | 2022-08-11 | 2022-08-11 | Polycrystalline silicon reduction furnace control method based on time series prediction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115032891B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115583654A (en) * | 2022-10-18 | 2023-01-10 | 科大智能物联技术股份有限公司 | Polysilicon reduction furnace current control method based on simulation learning |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE4416317A1 (en) * | 1993-05-17 | 1994-11-24 | Siemens Ag | Method and control device for controlling a material-processing process |
US6110214A (en) * | 1996-05-03 | 2000-08-29 | Aspen Technology, Inc. | Analyzer for modeling and optimizing maintenance operations |
CN104155956A (en) * | 2014-08-19 | 2014-11-19 | 东华大学 | Wireless temperature remote monitoring system based on Wi-Fi |
CN109034497A (en) * | 2018-08-31 | 2018-12-18 | 广东工业大学 | Prediction technique, system, medium and the equipment of polycrystalline reduction process energy consumption value |
CN109034275A (en) * | 2018-08-31 | 2018-12-18 | 广东工业大学 | Prediction technique, system, medium and the equipment of polycrystalline reduction process energy consumption value |
CN110858062A (en) * | 2018-08-22 | 2020-03-03 | 阿里巴巴集团控股有限公司 | Target optimization parameter obtaining method and model training method and device |
CN111273623A (en) * | 2020-02-25 | 2020-06-12 | 电子科技大学 | Fault diagnosis method based on Stacked LSTM |
CN111651954A (en) * | 2020-06-10 | 2020-09-11 | 嘉兴市像景智能装备有限公司 | Method for three-dimensional reconstruction of SMT electronic component based on deep learning |
CN111832839A (en) * | 2020-07-24 | 2020-10-27 | 河北工业大学 | Energy consumption prediction method based on sufficient incremental learning |
CN112488295A (en) * | 2020-11-23 | 2021-03-12 | 江苏科技大学 | Method for optimizing storage life prediction of LSTM network relay by cross validation algorithm |
CN112949836A (en) * | 2021-04-08 | 2021-06-11 | 北京航空航天大学 | Method for carrying out regression prediction on-line migration learning on time-varying distribution data |
CN113359212A (en) * | 2021-06-22 | 2021-09-07 | 中国石油天然气股份有限公司 | Reservoir characteristic prediction method and model based on deep learning |
CN114330647A (en) * | 2021-12-09 | 2022-04-12 | 浙江中控技术股份有限公司 | Model training method and device and silicon rod weight prediction method |
CN114610093A (en) * | 2022-03-09 | 2022-06-10 | 科大智能物联技术股份有限公司 | Burning furnace air supply control method based on variable period prediction of hot blast stove |
CN114655958A (en) * | 2020-12-22 | 2022-06-24 | 新疆新特晶体硅高科技有限公司 | Control method and system for polycrystalline silicon production |
-
2022
- 2022-08-11 CN CN202210958110.3A patent/CN115032891B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE4416317A1 (en) * | 1993-05-17 | 1994-11-24 | Siemens Ag | Method and control device for controlling a material-processing process |
US6110214A (en) * | 1996-05-03 | 2000-08-29 | Aspen Technology, Inc. | Analyzer for modeling and optimizing maintenance operations |
CN104155956A (en) * | 2014-08-19 | 2014-11-19 | 东华大学 | Wireless temperature remote monitoring system based on Wi-Fi |
CN110858062A (en) * | 2018-08-22 | 2020-03-03 | 阿里巴巴集团控股有限公司 | Target optimization parameter obtaining method and model training method and device |
CN109034497A (en) * | 2018-08-31 | 2018-12-18 | 广东工业大学 | Prediction technique, system, medium and the equipment of polycrystalline reduction process energy consumption value |
CN109034275A (en) * | 2018-08-31 | 2018-12-18 | 广东工业大学 | Prediction technique, system, medium and the equipment of polycrystalline reduction process energy consumption value |
CN111273623A (en) * | 2020-02-25 | 2020-06-12 | 电子科技大学 | Fault diagnosis method based on Stacked LSTM |
CN111651954A (en) * | 2020-06-10 | 2020-09-11 | 嘉兴市像景智能装备有限公司 | Method for three-dimensional reconstruction of SMT electronic component based on deep learning |
CN111832839A (en) * | 2020-07-24 | 2020-10-27 | 河北工业大学 | Energy consumption prediction method based on sufficient incremental learning |
CN112488295A (en) * | 2020-11-23 | 2021-03-12 | 江苏科技大学 | Method for optimizing storage life prediction of LSTM network relay by cross validation algorithm |
CN114655958A (en) * | 2020-12-22 | 2022-06-24 | 新疆新特晶体硅高科技有限公司 | Control method and system for polycrystalline silicon production |
CN112949836A (en) * | 2021-04-08 | 2021-06-11 | 北京航空航天大学 | Method for carrying out regression prediction on-line migration learning on time-varying distribution data |
CN113359212A (en) * | 2021-06-22 | 2021-09-07 | 中国石油天然气股份有限公司 | Reservoir characteristic prediction method and model based on deep learning |
CN114330647A (en) * | 2021-12-09 | 2022-04-12 | 浙江中控技术股份有限公司 | Model training method and device and silicon rod weight prediction method |
CN114610093A (en) * | 2022-03-09 | 2022-06-10 | 科大智能物联技术股份有限公司 | Burning furnace air supply control method based on variable period prediction of hot blast stove |
Non-Patent Citations (4)
Title |
---|
BANGWEN XI: "Time Series Forecasting in a CVD Reactor for Polysilicon Production", 《2021 IEEE 17TH INTERNATIONAL CONFERENCE ON AUTOMATION SCIENCE AND ENGINEERING》 * |
MA RUIHAO: "Application of a Combined Prediction Method Based on Temporal Decomposition and Convolutional Neural Networks for the Prediction of Consumption in Polysilicon Reduction Furnaces", 《PROCESSES》 * |
SOONSUNG HWANG: "A Novel Time Series based Seq2Seq Model for Temperature Prediction in Firing Furnace Process", 《PROCEDIA COMPUTER SCIENCE》 * |
郭久俊: "基于LSTM-Adaboost的多晶硅生产的能耗预测", 《计算机应用与软件》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115583654A (en) * | 2022-10-18 | 2023-01-10 | 科大智能物联技术股份有限公司 | Polysilicon reduction furnace current control method based on simulation learning |
CN115583654B (en) * | 2022-10-18 | 2023-11-21 | 科大智能物联技术股份有限公司 | Polysilicon reduction furnace current control method based on imitation learning |
Also Published As
Publication number | Publication date |
---|---|
CN115032891B (en) | 2022-11-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108134114B (en) | Proton exchange membrane fuel cell temperature control method | |
CN115032891B (en) | Polycrystalline silicon reduction furnace control method based on time series prediction | |
WO1999026118A1 (en) | Automatic control system for unit operation | |
US20200239982A1 (en) | Prediction Control Method And System For Component Contents In Rare Earth Extraction Process | |
CN109543916A (en) | Silicon rod growth rate prediction model in a kind of polycrystalline silicon reducing furnace | |
CN114510092B (en) | Transition packet internal temperature control system and method based on fuzzy PID (proportion integration differentiation) of prediction variable universe | |
CN115224698B (en) | Reactive power-voltage optimization control method for new energy power system based on digital twin | |
CN113862728B (en) | Pressure control method, system, equipment and medium for PEM pure water electrolysis hydrogen production | |
CN112947672A (en) | Maximum power point tracking method and device for photovoltaic cell | |
CN115441447A (en) | New energy power generation power prediction method | |
CN110673482A (en) | Power station coal-fired boiler intelligent control method and system based on neural network prediction | |
CN108388218B (en) | Correction self-adaptive batch process optimization method based on latent variable process migration model | |
CN101413115B (en) | Method for plasma assisted film deposition | |
CN113419432B (en) | Sewage treatment system accurate dosing method based on dynamic matrix control algorithm | |
JP3843999B2 (en) | Method for producing electrolysis product | |
CN113772674B (en) | Control method of polysilicon production reduction furnace | |
CN110912185A (en) | Design method of PID controller of automatic power generation control system of power grid containing wind power generation | |
CN108563913B (en) | Method and system for predicting energy consumption of reduction process of polycrystalline silicon reduction furnace | |
CN106355297A (en) | Electrolytic aluminum load characteristic-based power grid loss reduction optimization method | |
CN111998423A (en) | Electric heat storage circulating air quantity-water temperature control system and prediction control method thereof | |
CN115583654B (en) | Polysilicon reduction furnace current control method based on imitation learning | |
CN102108494B (en) | Deposition method for microcrystalline silicon thin films and device for monitoring plasma assisted deposition | |
CN113536685A (en) | Neural network extrapolation-based wind speed probability model modeling method | |
CN117421708B (en) | Alkaline electrolyzed water hydrogen reduction state monitoring method based on improved Gaussian mixture model | |
CN113241781A (en) | Photovoltaic fluctuation stabilizing method based on least square method and low-pass filtering |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |