CN102622367B - Method for filtering and compressing process data - Google Patents
Method for filtering and compressing process data Download PDFInfo
- Publication number
- CN102622367B CN102622367B CN201110032679.9A CN201110032679A CN102622367B CN 102622367 B CN102622367 B CN 102622367B CN 201110032679 A CN201110032679 A CN 201110032679A CN 102622367 B CN102622367 B CN 102622367B
- Authority
- CN
- China
- Prior art keywords
- data
- new data
- kmin
- kmax
- slope
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a method for filtering and compressing process data, which comprises the following steps: temporarily storing maximum gradient Kmax, minimum gradient Kmin, maximum gradient data UKmax and minimum gradient data UKmin related to a data source and last stored data related to the data source; receiving new data V related to the data source; using the maximum gradient data UKmax and the minimum gradient data UKmin to respectively calculate reference gradient K3 and reference gradient K4; calculating the gradient K of the new data; judging whether the gradient K is located in an interval [K3, K4] or not, if so, the new data are not stored, whether the K is included in an interval [Kmin, Kmax] or not is further judged, of K>Kmax, Kmin=K, and UKmax=V is updated; if the K is not located in the interval [K3, K4], the new data are stored and serve as the last stored data. The method uses a comparison gradient method to perform filtering and compressing judgment and enables the scale of temporarily stored data to be small and constant, and the complexity of the filter and compression algorithm is reduced remarkably.
Description
Technical field
The present invention relates to RTDB in Industry Control, especially relate to a kind of filtration and compression method of flow data.
Background technology
Real-time data base (Real Time DataBase, RTDB) for the service data of harvester, grasp the operation conditions of device, and the critical data of production run is monitored and analyzed, the problem occurring is processed in time, historical data is analysed scientifically, make production run state steady, material supply balance, reduces unit consumption, increase economic efficiency, reduce costs.
Minority u s company is monopolizing RTDB in Industry Control field at present.Their valuable product, conventionally only for large-scale huge enterprise, the needed real-time data base products quotation of medium-sized enterprise is Renminbi up to a million possibly, the deployment cost of great number becomes process control automation threshold together with informationalized, has restricted the development of domestic medium and small sized enterprises.
Data compression method is one of core methed of real-time data base.Current most popular real-time data compression method is " revolving door " compression method that the PI of U.S. OSI Software company (Plant Information System) system adopts, and this is a kind of outstanding method that can compress at high proportion sequential industrial flow data.As a kind of lossy compression method method, " revolving door " compression method is particularly suitable for the industrial flow data of sequential to compress (lower to general data compressibility), has the advantages that compressibility is high; Do not affect the resolution of data, reduction historical data that can be comparatively loyal during decompress(ion) simultaneously.
" but revolving door " method still has following obvious shortcoming:
(1) cannot determine in advance the scale of " temporary data set ", if there are a large amount of packed datas that do not need, can cause great waste to internal memory.
(2) along with the increase of " temporary data set " scale, the time overhead of certain step of method will become geometric growth, become system bottleneck.Extremely short when the device data sampling period, in the stable situation of equipment state, the performance of the method will decline to a great extent.To daily use, bring many problems.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of filtration and compression method of flow data, by the defect that solves existing revolving door algorithm existence.
The present invention solves the problems of the technologies described above data filtering and the compression method that the technical scheme adopting is a kind of real-time data base of proposition, comprises the following steps:
A. interim storage is about maximum slope Kmax, minimum slope Kmin, the maximum slope data U of a data source
kmaxwith a minimum slope data U
kmin, and about the upper memory point of these data;
B. receive the new data V about this data source;
C. use these maximum slope data U
kmaxwith minimum slope data U
kmincalculate respectively two with reference to slope K 3, K4;
D. calculate the slope K of new data;
E. judge whether slope K drops on interval [K3, K4] in, if it is do not store this new data, but further judge whether K is contained in [Kmin, Kmax], if Kmin > K makes Kmin=K and upgrades Ukmin=V, if K > Kmax makes Kmax=K and upgrades Ukmax=V; And
If in f. K does not drop on [K3, K4], store this new data, and using this new data as upper storage data;
K=(V-M)/Δ t1 wherein;
K3＝(U
Kmax-(M+ΔC))/Δt2；
K4＝(U
Kmin-(M-ΔC))/Δt3；
Wherein M is the upper storage data about this data source, and Δ C is compression deviation, Δ t1, and Δ t2, Δ t3 is two time intervals between data to be calculated.
In one embodiment of this invention, after above-mentioned steps b, also comprise: judge whether this new data is trust data, if yes then enter step c, otherwise abandons this data, returns to step a.
In one embodiment of this invention, after step b, also comprise: judge whether to receive this new data, if so, enter step c, otherwise, store this new data temporarily, as upper storage data, then return to step b.
In one embodiment of this invention, after step b, also comprise: judge whether maximum slope Kmax, minimum slope Kmin are empty, if so, utilize storage data on this new data and this to calculate maximum slope Kmax, minimum slope Kmin, then return to step b.
The step of in one embodiment of this invention, storing this new data in step f comprises: this new data is stored in a historical data base.
In one embodiment of this invention, the step of storing this new data in step f comprises: new data is stored in a buffer memory, and when meeting a trigger condition, trigger this buffer memory and file in a historical data base.
In one embodiment of this invention, before step a, also comprise one first filtering process, this first filtering process comprises: whether the data source that judges this new data available, the scan attribute of data source, the quality of data of the data type of new data, new data at least partly, to determine whether receive this new data.
In one embodiment of this invention, when the data type of this new data is discrete type, do not receive this new data.
In one embodiment of this invention, before step a, also comprise one second filtering process, this second filtering process comprises: according to the value of the data of the value of this new data and reception last time, and the time of reception of the data of the time of reception of this new data and reception last time, determine whether to receive this new data.
The present invention, owing to adopting above technical scheme, makes it compared with prior art, by comparing Slope Method, carrys out filtering data, thereby only needs interim storage about the maximum slope data U of data
kmax, minimum slope data U
kmin, and upper one storage data, greatly dwindled the scale of ephemeral data, and the judgement that whether data is met to storage demand is optimized, reducing slope calculates and number of times relatively, thereby filtration pressure compression algorithm complexity is significantly reduced, so the present invention compare existing revolving door algorithm and have higher operating efficiency.
Accompanying drawing explanation
For above-mentioned purpose of the present invention, feature and advantage can be become apparent, below in conjunction with accompanying drawing, the specific embodiment of the present invention is elaborated, wherein:
Fig. 1 illustrates the flow chart of data processing of the RTDB in Industry Control of one embodiment of the invention.
Fig. 2 illustrates the data filtering flow process of the RTDB in Industry Control of one embodiment of the invention.
Fig. 3 illustrates the data compression flow process of the RTDB in Industry Control of one embodiment of the invention.
Fig. 4 illustrates the data compression flow process of the RTDB in Industry Control of another embodiment of the present invention.
Fig. 5 A-5C illustrates the compression algorithm principle of one embodiment of the invention.
Embodiment
Fig. 1 illustrates the flow chart of data processing of RTDB in Industry Control according to an embodiment of the invention.Shown in Fig. 1, flow process is carried out in the system being comprised of data source 10, some attribute filtrator 20, event filter 30, level cache structure 40, compression filter 50, L2 cache structure 60 and historical data base 70.When using computing machine as data processing platform (DPP), level cache structure 40 and L2 cache structure 60 are general is to build in the such volatile memory of calculator memory, and 70 of historical data bases are to be structured in the hard disk of computing machine.
The data inlet point attribute filtrator 20 and the event filter 30 that from data source 10, take out, according to some default filtration according to judging that whether data are trust data, deposit trust data in level cache structure 40, and abandon untrusted data.
Data in level cache structure 40 will further be filtered through compression filter 50, and significant data is deposited in L2 cache structure 60.Compression filter 50 also can empty replacement to level cache structure 40.Data in L2 cache structured data, through filing, become a plurality of historical datas, after merging, are kept at enduringly in historical data base 70.
But in practical application, the packet of hardware collection is containing a large amount of noises, and the state of data source is also unpredictable, and these panoramic factors all affect the waveform of former data, if effectively do not controlled, can cause serious impact to efficiency of algorithm.Therefore,, before data enter compression process, must carry out suitable filtration.Data enter before compression filter 50, need through some attribute filtrator 20 and the two-layer filtrator of event filter 30.
Fig. 2 illustrates the data filtering flow process of the RTDB in Industry Control of one embodiment of the invention.Shown in Fig. 2, the flow process in empty frame is above the flow process that an attribute filtrator 20 is carried out, and the flow process in empty frame is below the flow process that event filter 30 is carried out.Although described in order the flow process of some attribute filtrator 20 and event filter 30 in a flow process, be appreciated that the execution that these flow processs can be separate.
From the data of data source output, first judge whether manual input (step S10), if so, directly arrive L2 cache structure (step S11), otherwise, inlet point attribute filtrator.
The task of some attribute filtrator is, according to the attribute of data source and state at that time, whether determination data receives.The following flow process of main execution: at step S12, judge that whether data source available: when data source state is not suitable for work (if data source is in off working state, or data source can not provide the situations such as accurate data in extreme environment), this data source data will not receive (step S13).At step S14, the scan attribute of judgement data collector, when the scan attribute of data source (in this time interval, data source data is possible meaningless) data when closing will not receive (step S13).At step S15 and S16, judge whether data are Boolean type, and the curve of Boolean type data is discrete, can not adopt flow data compress mode to process, therefore will not receive (step S13).At step S17, carry out quality of data filtration, the data that do not meet the demands for quality will not receive (data type and the definition that as data source, give are not inconsistent, and data are not medium in predefined value space), as step S18.Meet the data that other do not enter compression process condition, will not receive (step S19).Be appreciated that in different embodiment, can carry out the combination of the part steps in these steps, filter.
The task of event filter is that, according to the predefine of the attribute of data self and data source, whether determination data receives.Main carry out following flow process: at step S20, when the time tag that newly obtains data is during early than the time tag of upper storage data, do not filter, directly at step S11, receive data.Step S21 value event filtering: the difference of the data that new data and last time receive is greater than the abnormal variation value ExcDec of regulation, and new data and mistiming of receiving data last time are more than or equal to the minimum time value Excmin of regulation, received, delivered to compression filter (step S22).At step S23, carry out time-event filtration: the difference of the temporal information of the data that the temporal information of new data and last time receive is more than or equal to the maximum time value ExcMax of regulation, is received, and delivers to compression filter (step S22).If receive new data and last time the mistiming of data while being less than minimum time value, new data will not receive (step S24).
It is abnormal variation (for value event) and minimax time (for temporal filtering) that event filter relates to main parameter.
Through attribute filter with abnormal filtration after data, raw data noise rate reduces (although can not determine eliminate noise completely) greatly relatively, for the normal work of compression filter 50, plays vital effect.
At this, the data compression algorithm in compression filter 40, is that a plurality of data that same data source is continued are in time screened, and judges which data needs to preserve, and which data can abandon.This algorithm carries out the key of data filtering and compression.As the comparison with the embodiment of the present invention, the principle of the swinging door compression algorithm of prior art is first briefly described.In existing " revolving door " data compression, system often receives a new data, and all can go out one with the data configuration of nearest storage has two limits with the vertical parallelogram of X-axis.At this moment, if at new data with stored between data and exist a data V to drop on outside parallelogram, store the last data of new data, otherwise just continue to receive new data, without any data, be stored.Wherein, the length of side of parallelogram vertical direction is the twice of compression deviation amount, and compression deviation amount is set by practical application request, and each signal source can have the compression deviation amount of oneself.With a temporary data set, preserve the total data between stored data and new data, these data are kept in the internal memory of computing machine as important ephemeral data.
The defect that revolving door algorithm exists is in actual applications: first, the amount of capacity of temporary data set is unknown, if exist the data of a large amount of needs compression in flow data sequence, the unpredictable increase of gathering of that ephemeral data, causes internal memory waste even to overflow.Secondly, along with the increase of temporary data set, in algorithm, test the computing cost whether concentrated data of ephemeral data drop on this step in parallelogram and just become geometric growth, become the bottleneck of throughput of system.
Embodiments of the invention remain the theory based on " removing the not obvious data item of adjacent variation ", but take improved data compression method.An advantage of the method is can solve in " revolving door " compression algorithm flow process, the problem that temporary data set is excessive.
According to one embodiment of the invention, propose " slope comparison method " and replace temporary data set method.Fig. 4 A-4C illustrates the compression algorithm principle of one embodiment of the invention.In Fig. 4 A-4C, A point is last savepoint, and D point is latest data point, and B point is the point of slope maximum between A point and D point, and C point is the point of slope minimum between A point and D point.
In " slope comparison method " of one embodiment of the invention, only store the minimum and maximum data of slope temporarily.If the slope of new data and a upper storage data formation, between minimum and maximum slope, need not carry out other tests; If the slope of new data and a upper storage data formation is outside minimum and maximum slope, whether the data that only need to test maximum slope and minimum slope drop in the parallelogram of having stored data and new data formation, thereby determine storage current data (if it is outer to drop on parallelogram) or continue the new data of acceptance (if dropping in parallelogram).
For realizing above-mentioned method, build an ephemeral data buffer memory.This ephemeral data buffer memory can be structured in the secondary data buffer structure 50 shown in Fig. 1.At this, each observed reading is the single data in the individual data source that obtains of single pass.The cache contents of ephemeral data buffer memory comprises:
(1) the nearest storage numerical value of each corresponding point of buffer memory and upper one is observed numerical value, and these two values are open to user and other assemblies.
(2) the maximum slope Kmax of each data source of buffer memory, minimum slope Kmin, and record obtains the observation data U of Kmax and Kmin
kmax, U
kmin.
Fig. 2 illustrates the data compression method flow process of the RTDB in Industry Control of one embodiment of the invention.The flow process of slope comparison method is described below in conjunction with Fig. 2:
First at step S101, accept data;
At step S102, via an event filter 10, carry out event filtering, if insincere data abandon data at step S103, otherwise enter step S104.
At step S104, judge whether to receive these data, if not, show it is to receive for the first time this data, at step S105, these data exist in an ephemeral data buffer memory, then return to step S101.If so, enter step S106, judge whether maximum slope Kmax and minimum slope Kmin are empty, empty if, at step S107, utilize this point and a upper memory point to calculate maximum slope Kmax and minimum slope Kmin, then return to step S101.
At step S108, utilize minimum and maximum slope data U
kmax, U
kmincalculating K 3, K4, computing method are as follows:
K3＝(U
Kmax-(M+ΔC))/Δt2；
K4＝(U
Kmin-(M-ΔC))/Δt3。
Wherein M is upper storage data, and Δ C is compression deviation, Δ t2, and Δ t3 is two data U to be calculated
kmaxwith Δ C, and U
kminand the time interval between Δ C.
At step S109, calculate the slope K of current data.
K＝(V-M)/Δt1。
Wherein Δ t1 is the time interval between V and M, and V is current data.
At step S110, if the slope K of current point meets: K3≤K≤K4, judges whether K is contained in [Kmin, Kmax], if Kmin > K makes Kmin=K upgrade U simultaneously
kmin=V, if K > Kmax makes Kmax=K and upgrades U
kmax=V.
At step S111, current data does not meet preservation condition, does not preserve current data.But be that data observation needs, current data is write on one in an observed reading register, for other programs or flow process.
On the contrary, if in the slope K of step S112 current data not in [K3, K4] interval, preserve current data to L2 cache, upgrade up-to-date storage data simultaneously.
At step S113, judge whether to trigger L2 cache filing, if so, at step S114, file.
In any case flow process is processed next data at step S115, turn back to step S101.
In an embodiment of the present invention, the condition that triggers filing can be: time interval arrives, and L2 cache structure is full, system generation great change etc.
Variation example as the present embodiment, in Fig. 3, at step S112 ', directly stores current data in one historical data base, then enters step S115.
When the benefit of new algorithm is new data arrival, judge the choice of data by comparing slope, only retain the value of two minimum and maximum data of slope, buffer size is constant, can not cause internal memory waste and overflow.The time series of industrial real-time data has certain waveform rule (as sine wave) conventionally, experience shows that the compressibility of industrial real-time data is all higher, such data characteristic has determined that temporary data set generally has certain scale, and the decision operation efficiency that the compare operation of slope is compared on certain scale temporary data set is so higher.Compression experiment for mass data has also proved that the performance of new algorithm has advantage, thereby embodiments of the invention have more successfully solved the left problem of tradition " revolving door " compression algorithm.
Various embodiment described herein can be implemented in the computer-readable medium of for example combination of computer software, hardware or computer software and hardware.For hardware implementation, embodiment described herein can be at one or more special ICs (ASIC), digital signal processor (DSP), digital signal processor (DAPD), programmable logic device (PLD) (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, for carrying out other electronic installation of above-mentioned functions or the Selection and Constitute of said apparatus, implemented.In part situation, this class embodiment can implement by controller.
For implement software, embodiment described herein can by such as program module (procedures) and function module (functions) etc. independently software module implemented, wherein each module is carried out one or more functions described herein and operation.Software code can be implemented by the application software of writing in suitable programming language, can be stored in internal memory, by controller or processor, is carried out.
Although the present invention discloses as above with preferred embodiment; so it is not in order to limit the present invention, any those skilled in the art, without departing from the spirit and scope of the present invention; when doing a little modification and perfect, so protection scope of the present invention is worked as with being as the criterion that claims were defined.
Claims (9)
1. the filtration of flow data and a compression method, comprise the following steps:
A. interim storage is about maximum slope Kmax, minimum slope Kmin, the maximum slope data U of a data source
kmaxwith a minimum slope data U
kmin, and about this data source upper one storage data;
B. receive the new data V about this data source;
C. use these maximum slope data U
kmaxwith minimum slope data U
kmincalculate respectively two with reference to slope K 3, K4;
D. calculate the slope K of new data;
E. judge whether slope K drops on interval [K3, K4] in, if it is do not store this new data, but further judge whether K is contained in [Kmin, Kmax], if Kmin>K makes Kmin=K and upgrades Ukmin=V, if K>Kmax makes Kmax=K and upgrades Ukmax=V; And
If in f. K does not drop on [K3, K4], store this new data, and using this new data as upper storage data;
K=(V-M)/Δ t1 wherein;
K3=(U
Kmax-(M+ΔC))/Δt2；
K4=(U
Kmin-(M-ΔC))/Δt3；
Wherein M is the upper storage data about this data source, and Δ C is compression deviation, Δ t1, and Δ t2, Δ t3 is two time intervals between data to be calculated.
2. the method for claim 1, is characterized in that, after step b, also comprises:
Judge whether this new data is trust data, if yes then enter step c, otherwise abandons this data, returns to step a.
3. the method for claim 1, is characterized in that, after step b, also comprises:
Judge whether to receive this new data, if so, enter step c, otherwise, store this new data temporarily, as upper storage data, then return to step b.
4. the method for claim 1, is characterized in that, after step b, also comprises:
Judge that whether maximum slope Kmax, minimum slope Kmin are empty, if so, utilize this new data and on this storage data calculate maximum slope Kmax, minimum slope Kmin, then return to step b.
5. the method for claim 1, is characterized in that, the step of storing this new data in step f comprises: this new data is stored in a historical data base.
6. the method for claim 1, is characterized in that, the step of storing this new data in step f comprises: new data is stored in a buffer memory, and when meeting a trigger condition, trigger this buffer memory and file in a historical data base.
7. the method for claim 1, it is characterized in that, before step a, also comprise one first filtering process, this first filtering process comprises: whether the data source that judges this new data available, the scan attribute of data source, the quality of data of the data type of new data, new data at least partly, to determine whether receive this new data.
8. method as claimed in claim 7, is characterized in that, when the data type of this new data is discrete type, does not receive this new data.
9. the method as described in claim 1 or 7, it is characterized in that, before step a, also comprise one second filtering process, this second filtering process comprises: according to the value of the data of the value of this new data and reception last time, and the time of reception of the data of the time of reception of this new data and reception last time, determine whether to receive this new data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110032679.9A CN102622367B (en) | 2011-01-30 | 2011-01-30 | Method for filtering and compressing process data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110032679.9A CN102622367B (en) | 2011-01-30 | 2011-01-30 | Method for filtering and compressing process data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102622367A CN102622367A (en) | 2012-08-01 |
| CN102622367B true CN102622367B (en) | 2014-08-20 |
Family
ID=46562290
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110032679.9A Active CN102622367B (en) | 2011-01-30 | 2011-01-30 | Method for filtering and compressing process data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102622367B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103207887A (en) * | 2013-01-14 | 2013-07-17 | 广州佳都信息技术研发有限公司 | Efficient historical trend data storage method |
| CN104484673B (en) * | 2014-12-05 | 2017-10-20 | 南京大学 | The Supplementing Data method of real-time stream application of pattern recognition |
| CN104682962A (en) * | 2015-02-09 | 2015-06-03 | 南京邦耀科技发展有限公司 | Compression method for massive fuel gas data |
| CN112182034A (en) * | 2019-07-03 | 2021-01-05 | 河南许继仪表有限公司 | Data compression method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1459743A (en) * | 2002-05-24 | 2003-12-03 | 中国科学院软件研究所 | Self adapting history data compression method |
| CN101069353A (en) * | 2004-09-28 | 2007-11-07 | 通用电气公司 | Industrial data compression systems and methods |
| CN101923569A (en) * | 2010-07-09 | 2010-12-22 | 南京朗坤软件有限公司 | Storage method of structure type data of real-time database |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7228323B2 (en) * | 2003-06-10 | 2007-06-05 | Siemens Aktiengesellschaft | HIS data compression |
-
2011
- 2011-01-30 CN CN201110032679.9A patent/CN102622367B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1459743A (en) * | 2002-05-24 | 2003-12-03 | 中国科学院软件研究所 | Self adapting history data compression method |
| CN101069353A (en) * | 2004-09-28 | 2007-11-07 | 通用电气公司 | Industrial data compression systems and methods |
| CN101923569A (en) * | 2010-07-09 | 2010-12-22 | 南京朗坤软件有限公司 | Storage method of structure type data of real-time database |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102622367A (en) | 2012-08-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN201993755U (en) | Data filtration, compression and storage system of real-time database | |
| US11392582B2 (en) | Automatic partitioning | |
| US10169377B2 (en) | Managing a database management system using a set of stream computing data | |
| US10049136B1 (en) | Managing a stream computing environment using a projected database object | |
| CN105593818A (en) | Apparatus and method for scheduling distributed workflow tasks | |
| CN102622367B (en) | Method for filtering and compressing process data | |
| CN110334274A (en) | Information-pushing method, device, computer equipment and storage medium | |
| WO2017183065A1 (en) | Device and method for tuning relational database | |
| CN102999528B (en) | ETL tasks are offline in a kind of data warehouse and the method and device of data scrubbing | |
| CN103984726A (en) | Local revision method for database execution plan | |
| CN102750367A (en) | Big data checking system and method thereof on cloud platform | |
| CN111381970B (en) | Cluster task resource allocation method and device, computer device and storage medium | |
| CN103186603A (en) | Method, system and equipment for determining influence of SQL statements on performance of key businesses | |
| WO2021194673A1 (en) | Servicing queries during data ingress | |
| CN110825526B (en) | Distributed scheduling method and device based on ER relationship, equipment and storage medium | |
| CN106649718A (en) | Large data acquisition and processing method for PDM system | |
| CN117171157A (en) | Clearing data acquisition and cleaning method based on data analysis | |
| CN110597796B (en) | Big data real-time modeling method and system based on full life cycle | |
| CN102750353A (en) | Method for analyzing distributed data in key value library | |
| CN113220551A (en) | Index trend prediction and early warning method and device, electronic equipment and storage medium | |
| CN112150248B (en) | Method, system and device for counting hung goods amount based on batch flow fusion | |
| CN107807952A (en) | A kind of Apriori parallel methods, system and device based on Spark | |
| JP2004070633A (en) | Planning support system and planning support method | |
| CN116362212A (en) | Report generation method, device, equipment and storage medium | |
| Ravichandran et al. | A novel approach for optimal grouping of reusable software components for component based software development systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |