CN114186865A - Process industry energy consumption assessment and optimization method based on machine learning - Google Patents
Process industry energy consumption assessment and optimization method based on machine learning Download PDFInfo
- Publication number
- CN114186865A CN114186865A CN202111523414.9A CN202111523414A CN114186865A CN 114186865 A CN114186865 A CN 114186865A CN 202111523414 A CN202111523414 A CN 202111523414A CN 114186865 A CN114186865 A CN 114186865A
- Authority
- CN
- China
- Prior art keywords
- data
- energy consumption
- machine learning
- working condition
- model
- 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
- 238000005265 energy consumption Methods 0.000 title claims abstract description 39
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000010801 machine learning Methods 0.000 title claims abstract description 21
- 230000008569 process Effects 0.000 title claims abstract description 16
- 238000005457 optimization Methods 0.000 title claims abstract description 13
- 238000004519 manufacturing process Methods 0.000 claims abstract description 19
- 238000004140 cleaning Methods 0.000 claims abstract description 10
- 238000013210 evaluation model Methods 0.000 claims abstract description 5
- 238000011156 evaluation Methods 0.000 claims description 17
- 238000012549 training Methods 0.000 claims description 9
- 238000009826 distribution Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 6
- 238000009499 grossing Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 2
- 230000007246 mechanism Effects 0.000 claims description 2
- 230000001960 triggered effect Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 5
- 230000009467 reduction Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000004134 energy conservation Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 239000004568 cement Substances 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010926 purge Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
-
- 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
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
-
- 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
- Y02P90/30—Computing systems specially adapted for manufacturing
-
- 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
- Y02P90/80—Management or planning
- Y02P90/82—Energy audits or management systems therefor
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Manufacturing & Machinery (AREA)
- General Factory Administration (AREA)
Abstract
The invention relates to a process industry energy consumption assessment and optimization method based on machine learning, which is characterized by comprising the following steps of: s1: self-cleaning; s2: dividing the working conditions; s3: modeling; s4: self-updating; s5: evaluating; the invention can automatically download data without influencing normal production, automatically clean data based on machine learning, generate a productivity evaluation model according to working conditions, regularly update the model to be suitable for new production working conditions, evaluate the current energy consumption state in real time and output an adjustment suggestion.
Description
Technical Field
The invention relates to a process industry energy consumption evaluation and optimization method based on machine learning.
Background
Energy conservation and emission reduction are basic national policies in China and are indispensable ways for sustainable development; the process industry consumes about 70% of the energy of the whole industry, which shows that the process industry in China has great potential in energy conservation and consumption reduction, and the development of energy conservation and consumption reduction has important significance for society and ecological environment while reducing the cost and improving the benefit of enterprises, and is more in line with the development direction of economy in China.
The flow production process has obvious multiple working conditions and instability, the working conditions have different and unstable changes every time the control quantity is adjusted, and the multiple control quantities have coupling, so that the traditional data analysis method is difficult to eliminate the unstable working conditions.
At present, a plurality of products for saving energy and reducing consumption exist in the market, but most of the products are solid models or expert experiences, and dynamic flexible applicability to new stable working conditions and production modes is lacked.
Disclosure of Invention
In view of the current situation of the prior art, the technical problem to be solved by the present invention is to provide a process industry energy consumption evaluation and optimization method based on machine learning, which can automatically eliminate unstable working conditions and update at regular time to be suitable for new production working conditions, and further can flexibly and real-timely evaluate the current energy consumption state and output adjustment suggestions to greatly improve the applicability.
The technical scheme adopted by the invention for solving the technical problems is as follows: a process industry energy consumption assessment and optimization method based on machine learning is characterized by comprising the following steps:
s1: self-cleaning, cleaning different types of data using machine learning and feature configuration;
the S1 includes:
s11, eliminating shutdown and noise, and eliminating shutdown and noise data of each characteristic by using a method of combining a shielding measuring point with a denoising algorithm;
s12: mode division, distinguishing data of different production processes;
s13: quality judgment and feature screening are carried out, and unqualified quality data and excessive noise features in each mode are removed;
s2: dividing data into different working conditions mainly by using machine learning according to working conditions, and carrying out secondary cleaning on the data by using a multi-dimensional denoising algorithm;
the S2 includes:
s21: classifying the working condition characteristics in the mode data set according to working conditions in a clustering or expert experience mode and the like to obtain each working condition data set under different modes;
s22: carrying out multidimensional denoising, namely carrying out multidimensional density denoising on the machine hour and the energy consumption under each working condition to ensure that the data distribution is relatively centralized and stable;
s3: modeling, namely determining the optimal point of the productivity of the working condition according to the weight, and establishing an energy consumption evaluation model;
the S3 includes:
s31: optimizing the working condition, establishing an evaluation function based on the weight of the station hour and the energy consumption, and taking a sample with the highest evaluation value in the working condition data set as a working condition optimal point;
s32: establishing an evaluation model, determining a capacity relation model by machine learning or expert experience according to distribution of the station hour and energy consumption, and defining evaluation modes of different energy consumption levels on the basis of the relation model;
s4: self-updating, which is to selectively or regularly update the model parameters suitable for new working conditions;
the S4 includes:
s41: making a model updating triggering mechanism;
s42: training a model;
s5: evaluating, namely evaluating the data in real time based on the model;
the S5 includes:
s51: identifying the current data state, and identifying whether the current data is normal data in an operating state according to the normal range of the shielded measuring points and the characteristics;
s52: evaluating the energy consumption level;
s53: push adjustment suggestions.
Preferably, the step S11 of rejecting the stoppages and noise and rejecting the noise data of each feature by using a method of combining the masked measuring points and the denoising algorithm includes, but is not limited to, the following steps:
s111, eliminating the catastrophe points of each characteristic variable in the shielded measuring points by adopting median smoothing, and filling the catastrophe points with previous values, thereby effectively preventing false shutdown or false startup; then, obtaining an integral operation time period based on a starting-up threshold and a shielding rule set by a user or obtained by self-calculation so as to eliminate the shutdown time period data of all other variables;
and S112, on the basis of the step S111, performing combined denoising on each feature from the angles of physical threshold, variation amplitude, data distribution, data stability and the like.
Preferably, the pattern division in step S12 distinguishes data of different production processes, including but not limited to the following steps:
s121: dividing the data into mode data sets of different production processes based on the combined conditions of the production modes;
s122: on the basis of step S121, each feature range of the pattern data set is extracted.
Preferably, the quality judgment and feature screening in step S13 eliminates the unqualified quality data and the characteristics with excessive noise in each mode, including but not limited to the following steps:
s131: deleting unqualified quality data and samples thereof based on qualified conditions of each mode concerning quality;
s132: in step S131, the noise condition of each feature is determined, and features with a large amount of noise data are removed.
Preferably, the step S41 includes the following steps: (1) updating and triggering training of configuration information; (2) triggering training according to a model updating period; (3) multiple retrains triggered by a model training failure.
Preferably, the step S52 of evaluating the energy consumption level includes, but is not limited to, the following steps:
s521: on the basis of the step S52, judging the production mode and the working condition thereof, and calculating the normal energy consumption level under the same time;
s522: and outputting the energy consumption evaluation in the same time, the evaluation results of whether the lifting space and the quality are qualified or not in the relatively optimal state and the like on the basis of the step S521.
Preferably, the push adjustment suggestion of step S53 includes, but is not limited to, the following steps: s531: if the quality is qualified, pushing the adjustment direction of the controllable variable at the optimal point of the relative working condition on the basis of the step S522; if the quality is not qualified, the adjustment suggestion is not pushed.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention has self-adaptability, and can automatically learn the working condition of the system and immediately apply the system after being deployed to a new production system; and the model can be automatically updated at regular time to quickly adapt to the application of new working conditions.
2. The invention has a high-efficiency and stable data cleaning algorithm and can automatically clean abnormal data and unstable working conditions.
3. The invention can correlate the energy consumption analysis under quality to better fit the requirements of users.
4. The invention can output the operation directivity suggestion of energy saving and consumption reduction in real time.
Drawings
FIG. 1 is a block diagram of the framework of the present invention;
FIG. 2 is a diagram of the operating state of the present invention based on the shielded measurement point determination;
FIG. 3 is a purge schematic of manipulated variables of the present invention;
FIG. 4 is a schematic illustration of the purging of the controlled variable of the present invention;
FIG. 5 is a schematic illustration of the cleaning of the quality data of the present invention;
FIG. 6 is a schematic diagram of the clustering division of the present invention;
FIG. 7 is a schematic diagram of the behavior training results of the present invention;
fig. 8 is a schematic diagram of the energy consumption alarm evaluation result of the present invention.
Detailed Description
Unless defined otherwise, technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention belongs. The use of "first," "second," and similar terms in the present application do not denote any order, quantity, or importance, but rather the terms are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", and the like are used merely to indicate relative positional relationships, and when the absolute position of the object being described is changed, the relative positional relationships may also be changed accordingly.
To maintain the following description of the embodiments of the present invention clear and concise, a detailed description of known functions and known components of the invention have been omitted.
As shown in fig. 1, in this embodiment, a cement mill in a production line of a cement plant is taken as an example, and a process industry energy consumption evaluation and optimization method based on machine learning includes two steps of modeling and evaluation.
Modeling:
(1) the historical data of a certain time is automatically downloaded from the database, and the operation flag bit label is obtained based on the combination judgment of the shielding measuring points, as shown in fig. 2: a flag bit of 1 indicates an operating state, and 0 indicates shutdown.
(2) For each feature, the shutdown data of the feature is respectively eliminated, the abnormal data is eliminated by using the upper limit and the lower limit, and the data of different types are cleaned by using a denoising algorithm, wherein the denoising algorithm used in the embodiment of the scheme comprises the following steps: median denoising, variance denoising, density denoising and standard score denoising, wherein the algorithm content is as follows: median denoising, and effective fields in the configuration are as follows: kernel, e _ limit. The method is characterized in that a point with a large difference value before and after the median is smoothed is regarded as noise from the angle of continuous change of data, and a variable P is [ P ]1,p2,…,pi,…pn]Wherein p isiIs the ith value of the variable P, which is the variable after median smoothingWhereinIs calculated by taking the window length as kernel based on the formula (1)iCorresponding median (kernel is odd, denoted 2k + 1):
e _ limit is the maximum percentage of the amplitude of the variable P before and after smoothing, and the difference threshold limit is calculated according to the formula (2) [ < l >1,l2,…li,…ln]:
And (3) denoising the variance, wherein effective fields in the configuration are as follows: kernel, zscore. The method regards the point with large variance as noise from the angle of slow change of normal continuous data, and the variance corresponding to the variable P is S ═ S1,s2,…si,…sn]Wherein s isiP is calculated by taking the window length as kernel based on the formula (3)iThe corresponding variance (kernel is odd, denoted 2k + 1):
si=std([pi-k,…pi,…pi+k]) (3)
denoising the variance S as a standard score.
Density denoising (algorithm after improvement on DBSCAN), the configuration effective fields are: firstly, in order to facilitate the general eps adjustment of the variables with the same sparsity, the density radius eps and the neighborhood minimum sample size n are normalized to obtain M ═ M [ M ] by performing the minimum and maximum normalization on the variable P based on a formula (4)1,m2,…mi,…mn],。
Wherein m isiFor the ith value of variable M, min is a function to find the minimum value and max is a function to find the maximum value. Determining n nearest neighbor sample sets NS ═ NS of each sample based on Euclidean distance1,ns2,…nsi,…nsn]The samples whose maximum distance from the nearest neighbor sample set is smaller than the density radius eps are regarded as the neighborhood center data set, the center data set and the nearest neighbor samples thereof are regarded as normal data, and the other data are regarded as noise.
Standard score denoising, the configuration of the effective fields are: zscore. The method is characterized in that a point with large deviation from a mean value is regarded as noise from the angle of data distribution, and a standard value corresponding to a variable P is Z ═ Z1,z2,…zi,…zn]Wherein z isiIs calculated based on the formula (5)iMagnitude of deviation from mean:
zi=|pi-mean(P)|/std(P) (5)
zscore is the deviation threshold, and data greater than zscore in Z is considered noise.
Based on the algorithm, the cleaning effect is good: the manipulated variables are cleaned as shown in FIG. 3, the controlled variables are cleaned as shown in FIG. 4, and the quality data are cleaned as shown in FIG. 5, where the red data are the noise being cleaned. And aligning and merging the denoised data sets into an effective data set D.
(3) And performing mode division on the cleaned features according to the features of the production modes, and dividing the working conditions by using a clustering algorithm of automatic parameter adjustment in each mode, wherein label represents a clustering result and is also a working condition category as shown in fig. 6.
(4) For each working condition, dividing original data into three types of data based on a density denoising algorithm and the like: sparse data (blue color points), neighborhood center data (green color points), and center neighbor point data (orange color points), selecting a sample corresponding to the maximum value of the evaluation function from the neighborhood center data as an optimal point, and generating a yield-consumption relationship curve and an envelope curve according to a non-sparse data set as shown in fig. 7.
An evaluation step:
(1) and judging that the current state is normal data of operation.
(2) And judging that the production mode is the first working condition of semi-finished powder, the quality is qualified, and the lifting space under the output corresponding working condition is 17.8 percent, as shown in figure 8.
(3) The energy consumption assessment and adjustment recommendations are output as shown in fig. 8.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that various changes in the embodiments and modifications thereof may be made, and equivalents may be substituted for elements thereof; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. A process industry energy consumption assessment and optimization method based on machine learning is characterized by comprising the following steps:
s1: self-cleaning, cleaning different types of data using machine learning and feature configuration;
the S1 includes:
s11, eliminating shutdown and noise, and eliminating shutdown and noise data of each characteristic by using a method of combining a shielding measuring point with a denoising algorithm;
s12: mode division, distinguishing data of different production processes;
s13: quality judgment and feature screening are carried out, and unqualified quality data and excessive noise features in each mode are removed;
s2: dividing data into different working conditions mainly by using machine learning according to working conditions, and carrying out secondary cleaning on the data by using a multi-dimensional denoising algorithm;
the S2 includes:
s21: classifying the working condition characteristics in the mode data set according to working conditions in a clustering or expert experience mode and the like to obtain each working condition data set under different modes;
s22: carrying out multidimensional denoising, namely carrying out multidimensional density denoising on the machine hour and the energy consumption under each working condition to ensure that the data distribution is relatively centralized and stable;
s3: modeling, namely determining the optimal point of the productivity of the working condition according to the weight, and establishing an energy consumption evaluation model;
the S3 includes:
s31: optimizing the working condition, establishing an evaluation function based on the weight of the station hour and the energy consumption, and taking a sample with the highest evaluation value in the working condition data set as a working condition optimal point;
s32: establishing an evaluation model, determining a capacity relation model by machine learning or expert experience according to distribution of the station hour and energy consumption, and defining evaluation modes of different energy consumption levels on the basis of the relation model;
s4: self-updating, which is to selectively or regularly update the model parameters suitable for new working conditions;
the S4 includes:
s41: making a model updating triggering mechanism;
s42: training a model;
s5: evaluating, namely evaluating the data in real time based on the model;
the S5 includes:
s51: identifying the current data state, and identifying whether the current data is normal data in an operating state according to the normal range of the shielded measuring points and the characteristics;
s52: evaluating the energy consumption level;
s53: push adjustment suggestions.
2. The method for process industry energy consumption assessment and optimization based on machine learning of claim 1, wherein the step of S11 is to remove the stoppages and noise, and to remove the noise data of each feature by using a method of combining the shielded measuring points and the denoising algorithm, including but not limited to the following steps:
s111, eliminating the catastrophe points of each characteristic variable in the shielded measuring points by adopting median smoothing, and filling the catastrophe points with previous values, thereby effectively preventing false shutdown or false startup; then, obtaining an integral operation time period based on a starting-up threshold and a shielding rule set by a user or obtained by self-calculation so as to eliminate the shutdown time period data of all other variables;
and S112, on the basis of the step S111, performing combined denoising on each feature from the angles of physical threshold, variation amplitude, data distribution, data stability and the like.
3. The method for process industry energy consumption assessment and optimization based on machine learning of claim 1, wherein said pattern division in step S12 distinguishes data of different production processes, including but not limited to the following steps:
s121: dividing the data into mode data sets of different production processes based on the combined conditions of the production modes;
s122: on the basis of step S121, each feature range of the pattern data set is extracted.
4. The method for process industry energy consumption assessment and optimization based on machine learning of claim 1, wherein the quality judgment and feature screening in step S13 are performed to remove the unqualified quality data and the noisy features in each mode, including but not limited to the following steps:
s131: deleting unqualified quality data and samples thereof based on qualified conditions of each mode concerning quality;
s132: in step S131, the noise condition of each feature is determined, and features with a large amount of noise data are removed.
5. The method for evaluating and optimizing process industry energy consumption based on machine learning of claim 1, wherein the step S41 comprises the following steps: (1) updating and triggering training of configuration information; (2) triggering training according to a model updating period; (3) multiple retrains triggered by a model training failure.
6. The method for process industry energy consumption assessment and optimization based on machine learning of claim 1, wherein said assessing energy consumption level of step S52 includes but is not limited to the following steps:
s521: on the basis of the step S52, judging the production mode and the working condition thereof, and calculating the normal energy consumption level under the same time;
s522: and outputting the energy consumption evaluation in the same time, the evaluation results of whether the lifting space and the quality are qualified or not in the relatively optimal state and the like on the basis of the step S521.
7. The method for process industry energy consumption assessment and optimization based on machine learning according to claim 1, wherein said step S53 of pushing and adjusting recommendation includes but is not limited to the following steps:
s531: if the quality is qualified, pushing the adjustment direction of the controllable variable at the optimal point of the relative working condition on the basis of the step S522; if the quality is not qualified, the adjustment suggestion is not pushed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111523414.9A CN114186865A (en) | 2021-12-13 | 2021-12-13 | Process industry energy consumption assessment and optimization method based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111523414.9A CN114186865A (en) | 2021-12-13 | 2021-12-13 | Process industry energy consumption assessment and optimization method based on machine learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114186865A true CN114186865A (en) | 2022-03-15 |
Family
ID=80543575
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111523414.9A Pending CN114186865A (en) | 2021-12-13 | 2021-12-13 | Process industry energy consumption assessment and optimization method based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114186865A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024002026A1 (en) * | 2022-06-30 | 2024-01-04 | 中兴通讯股份有限公司 | Energy-consumption optimization method, system and apparatus, and storage medium |
-
2021
- 2021-12-13 CN CN202111523414.9A patent/CN114186865A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024002026A1 (en) * | 2022-06-30 | 2024-01-04 | 中兴通讯股份有限公司 | Energy-consumption optimization method, system and apparatus, and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Sun et al. | Multiclassification of tool wear with support vector machine by manufacturing loss consideration | |
WO2019237492A1 (en) | Semi-supervised learning-based abnormal electricity utilization user detection method | |
CN112819107B (en) | Artificial intelligence-based fault prediction method for gas pressure regulating equipment | |
CN105205111A (en) | System and method for mining failure modes of time series data | |
CN111401573B (en) | Working condition state modeling and model correcting method | |
CN114186865A (en) | Process industry energy consumption assessment and optimization method based on machine learning | |
CN113515512A (en) | Quality control and improvement method for industrial internet platform data | |
CN114417971A (en) | Electric power data abnormal value detection algorithm based on K nearest neighbor density peak clustering | |
CN113259379A (en) | Abnormal alarm identification method, device, server and storage medium based on incremental learning | |
US20220243347A1 (en) | Determination method and determination apparatus for conversion efficiency of hydrogen production by wind-solar hybrid electrolysis of water | |
CN117152119A (en) | Profile flaw visual detection method based on image processing | |
CN111177495A (en) | Method for intelligently identifying data content and generating corresponding industry report | |
CN109951499A (en) | A kind of method for detecting abnormality based on network structure feature | |
CN116738551B (en) | Intelligent processing method for acquired data of BIM model | |
CN112613233A (en) | Algorithm for discovering environmental monitoring abnormal data based on single-classification support vector machine model | |
Borisov et al. | Neuro-fuzzy cognitive temporal models for predicting multidimensional time series with fuzzy trends | |
CN110751202B (en) | Method for establishing load soft measurement model of cement combined grinding ball mill and application | |
CN117113233A (en) | Hierarchical energy structure scene construction method and energy consumption abnormal link tracing method | |
CN110866829A (en) | Stock price prediction method based on integrated learning model | |
CN112991093B (en) | Electric larceny detection method and system based on edge calculation | |
CN113222268B (en) | Multi-mode reasoning-based tobacco baking quality prediction model establishment method | |
CN111125082B (en) | Cement process parameter data analysis method based on association rule mining | |
CN114493234A (en) | Method for identifying key pressure control points of water supply pipe network | |
CN110457906B (en) | Intelligent alarm method for network security event | |
CN111652734B (en) | Financial information management system based on block chain and big data |
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 | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20230811 Address after: No. 320 and 436 Pingtang East Road, Puxing Street, Xinjin District, Chengdu City, Sichuan Province, 611400 Applicant after: Sinoma Intelligent Technology (Chengdu) Co.,Ltd. Address before: 303 Hanzhongmen street, Gulou District, Nanjing City, Jiangsu Province Applicant before: NANJING KISEN INTERNATIONAL ENGINEERING Co.,Ltd. |