CN109408545B - Target polymerization operation processing method and device - Google Patents
Target polymerization operation processing method and device Download PDFInfo
- Publication number
- CN109408545B CN109408545B CN201811158944.6A CN201811158944A CN109408545B CN 109408545 B CN109408545 B CN 109408545B CN 201811158944 A CN201811158944 A CN 201811158944A CN 109408545 B CN109408545 B CN 109408545B
- Authority
- CN
- China
- Prior art keywords
- converging
- sub
- sliding window
- target polymerization
- length
- 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
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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Abstract
The present invention provides a kind of target polymerization operation processing method and devices, wherein this method comprises: obtaining the parameter area of the sliding window length of target polymerization operation and accelerator；Determine whether the sliding window length of target polymerization operation exceeds parameter area；In the case where determining beyond parameter area, target polymerization operation is split as multiple sub- converging operations, wherein the sliding window length of every sub- converging operation is all located in parameter area；Multiple sub- converging operations are accelerated by accelerator, obtain the implementing result of multiple sub- converging operations；The implementing result of multiple sub- converging operations is merged, the implementing result of target polymerization operation is obtained.It is not high to solve existing accelerator data fitness through the above way, caused by the converging operation excessive to sliding window length the problem of can not handling, reach and promoted accelerator in the parameter adaptability of processing sliding window converging operation, to improve the technical effect of its performance for handling kernel function.
Description
Technical field
The present invention relates to Internet technical field, in particular to a kind of target polymerization operation processing method and device.
Background technique
In numerous applications, time series analysis is vital, is arrived with that can quickly respond based on Fast Persistence
The data flow reached.For example, multiple client monitors the price fluctuation of stock market in the application of stock market, this is then needed
A system can effectively response analysis be requested for different clients.For example, the analysis that each client may need
It is different (for example, volatility rate weekly, monthly stock price average etc.).In addition to financial field, in medical treatment & health, commercially determine
In the fields such as plan, scientific algorithm, social media and network-control, effective Data Stream Processing is also very important.
In data stream management system, then client is matched by registering analysis request in upcoming data flow
Window size (range) and sliding distance (slide) are set, to generate lasting analysis result.Wherein, sliding distance
(slide) indicate that the time span of update result, window size (range) indicate to need to carry out the size of data of converging operation.
For example, range is equal to 5 minutes in stock application configuration, slide is equal to 3 minutes, shows primary using needing to update for every 3 minutes
As a result, and result from 5 minutes data in the past.
For how effectively to handle converging operation, currently no effective solution has been proposed.
Summary of the invention
The embodiment of the invention provides a kind of target polymerization operation processing method and devices, are slided with promoting accelerator in processing
The parameter adaptability of dynamic window converging operation, to improve its performance for handling kernel function.
According to an aspect of the present invention, a kind of target polymerization operation processing method is provided, this method comprises:
Obtain the parameter area of the sliding window length of target polymerization operation and accelerator；
Determine whether the sliding window length of the target polymerization operation exceeds the parameter area；
In the case where determining beyond the parameter area, target polymerization operation is split as multiple sub- polymerizations and is grasped
Make, wherein the sliding window length of every sub- converging operation is all located in the parameter area；
The multiple sub- converging operation is accelerated by accelerator, obtains the implementing result of multiple sub- converging operations；
The implementing result of multiple sub- converging operations is merged, the implementing result of the target polymerization operation is obtained.
In one embodiment, the target polymerization is operated the step of being split as multiple sub- converging operations includes:
Obtain the upper limit value in the parameter area；
The target polymerization operation order is split as N number of sub- converging operation, wherein the cunning of preceding N-1 sub- converging operations
The length of dynamic window is the upper limit value, and the length of the sliding window of the sub- converging operation of n-th is less than or equal to the upper limit value.
In one embodiment, the implementing result of multiple sub- converging operations is merged, obtains the target polymerization
The step of implementing result of operation includes:
The merging that the amount of offsetting is carried out to the implementing result of the sub- converging operation obtains holding for the target polymerization operation
Row result, wherein the offset of every sub- converging operation is according to each sub- converging operation before being located at current sub- converging operation
Sliding window length determine.
In one embodiment, the target polymerization operation is one of the following: accumulation operations, repeated subtraction ask maximum
It is worth, minimize, averages.
In one embodiment, the accelerator is the dedicated accelerator of time series.
On the other hand, a kind of target polymerization operation processing device is provided, which includes:
Module is obtained, the parameter area of the sliding window length for obtaining target polymerization operation and accelerator；
Determining module, for determining whether the sliding window length of the target polymerization operation exceeds the parameter area；
Module is split, in the case where determining beyond the parameter area, target polymerization operation to be split as
Multiple sub- converging operations, wherein the sliding window length of every sub- converging operation is all located in the parameter area；
Accelerating module obtains multiple sub- polymerization behaviour for accelerating by accelerator to the multiple sub- converging operation
The implementing result of work；
Merging module is merged for the implementing result to multiple sub- converging operations, obtains the target polymerization operation
Implementing result.
In one embodiment, the fractionation module includes:
Acquiring unit, for obtaining the upper limit value in the parameter area；
Split cells, for the target polymerization operation order to be split as N number of sub- converging operation, wherein preceding N-1 son
The length of the sliding window of converging operation is the upper limit value, and the length of the sliding window of the sub- converging operation of n-th is less than or equal to
The upper limit value.
In one embodiment, the merging module is specifically used for having the implementing result of the sub- converging operation
The merging of offset obtains the implementing result of the target polymerization operation, wherein the offset of every sub- converging operation is according to position
The length of the sliding window of each sub- converging operation before current sub- converging operation determines.
Another aspect provides a kind of computer equipment, comprising: memory, processor and storage are on a memory and can
The computer program run on a processor, which is characterized in that the processor is realized above-mentioned when executing the computer program
The step of method.
Another aspect provides a kind of non-volatile computer readable storage medium storing program for executing, is stored thereon with computer program,
The step of being characterized in that, the above method realized when the computer program is executed by processor.
In embodiments of the present invention, the target polymerization of the parameter area by sliding window length beyond accelerator, which operates, splits
It is all located at the sub- converging operation in the parameter area for multiple sliding window length, then, then by accelerator to multiple sons
Converging operation is accelerated, and the implementing result of multiple sub- converging operations is obtained, finally to the implementing result of multiple sub- converging operations
It merges, obtains the implementing result of target polymerization operation.Solves existing accelerator data fitness through the above way
It is not high, caused by the converging operation excessive to sliding window length the problem of can not handling, reached promotion accelerator and located
The parameter adaptability of sliding window converging operation is managed, to improve the technical effect of its performance for handling kernel function.
It will be apparent to a skilled person that can be not limited to the objects and advantages that the present invention realizes above specific
It is described, and the above and other purpose that the present invention can be realized will be more clearly understood according to following detailed description.
And it is to be understood that aforementioned description substantially and subsequent detailed description are exemplary illustration and explanation, not
The limitation to the claimed content of the present invention should be used as.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, not
Constitute limitation of the invention.In the accompanying drawings:
Fig. 1 is the method flow diagram of target polymerization operation processing method according to an embodiment of the present invention；
Fig. 2 be accelerator according to an embodiment of the present invention can support window length be 8 schematic diagram；
The case where Fig. 3 is according to an embodiment of the present invention when length of window is 18, and accelerator cannot be supported schematic diagram；
Fig. 4 is the schematic diagram that sliding window converging operation according to an embodiment of the present invention decomposes sub-operation 1；
Fig. 5 is the schematic diagram that sliding window converging operation according to an embodiment of the present invention decomposes sub-operation 2；
Fig. 6 is the schematic diagram that sliding window converging operation according to an embodiment of the present invention decomposes sub-operation 3；
Fig. 7 is that sub- output sequence according to an embodiment of the present invention merges schematic diagram；
Fig. 8 is the configuration diagram of server according to an embodiment of the present invention；
Fig. 9 is the structural block diagram of target polymerization operation processing device according to an embodiment of the present invention.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, right below with reference to embodiment and attached drawing
The present invention is described in further details.Here, exemplary embodiment and its explanation of the invention is used to explain the present invention, but simultaneously
It is not as a limitation of the invention.
It should be noted that for purposes of clarity, unrelated to the invention, the common skill in this field is omitted in attached drawing and explanation
The expression and description of component known to art personnel and processing.
The feature for describing and/or showing for a kind of embodiment can be in a manner of same or similar one or more
It uses in a other embodiment, is combined with the feature in other embodiment, or the feature in substitution other embodiment.
It should be emphasized that term "comprises/comprising" refers to the presence of feature, element, step or component when using herein, but simultaneously
It is not excluded for the presence or additional of one or more other features, element, step or component.
It is numerous particularly with calculating in view of the bottleneck that converging operation is whole system is effectively treated in data stream management system
The financial field and scientific algorithm field of weight.Such as: maximizing minimizes, average value, asks the operations such as preceding N, is all polymerization
Operation.In order to accelerate to these converging operations, generally not only propose the reconstruction on some algorithms, also using FPGA,
The special chips such as ASIC accelerate sliding window to polymerize.
However, needing access history data in large quantities due to the build-in attribute of sliding window converging operation, this is to hardware-accelerated
Bring adaptability problem.Because in hardware realization, it is thus necessary to determine that the size of buffer area carrys out buffer history data, and sliding window
The size for the historical data that converging operation needs is determined by range and slide.Meanwhile the range of sliding window converging operation
It is entirely to be determined according to user demand with slide, is the variable that can not be determined.
Based on problem above, propose in this example a kind of method for realizing any length of window sliding window converging operation and
Device, to promote the dedicated accelerator of time series in the parameter adaptability of processing sliding window converging operation, to improve at its
Manage the performance of kernel function.
As shown in Figure 1, providing a kind of target polymerization operation processing method in this example, may include steps of:
Step 101: obtaining the parameter area of the sliding window length of target polymerization operation and accelerator；
Wherein, the parameter area of the sliding window length of accelerator can be the maximum sliding window that the accelerator can be supported
Mouth length.For example, parameter area is 0 to 10, then showing that the maximum sliding window length that the accelerator can be supported is 10.
Step 102: determining whether the sliding window length of the target polymerization operation exceeds the parameter area；
The sliding window length of above-mentioned target polymerization operation for example can be 5, can be 20, can be 15 etc..If
The sliding window length of target polymerization operation is 5, then just without departing from parameter area, if the sliding window of target polymerization operation
Length is 15 or 20, then just exceeding parameter area.The target polymerization for being 5 for sliding window length operates, and can directly lead to
It crosses current accelerator to be accelerated, the target polymerization for being 15 or 20 for sliding window length operates, it is necessary to the target
Converging operation is handled, and can just be accelerated based on current accelerator after processing.
Step 103: in the case where determining beyond the parameter area, target polymerization operation being split as multiple sons
Converging operation, wherein the sliding window length of every sub- converging operation is all located in the parameter area；
When realizing, then target polymerization operation order is split as by the upper limit value in available parameter area
N number of sub- converging operation, wherein the length of the sliding window of preceding N-1 sub- converging operations is the upper limit value, the polymerization of n-th
The length of the sliding window of operation is less than or equal to the upper limit value.For example, upper limit value is 8, the sliding window of target polymerization operation
Length be 20, then three sub- converging operations can be split as, length is respectively as follows: 8,8,4.
Step 104: the multiple sub- converging operation being accelerated by accelerator, obtains holding for multiple sub- converging operations
Row result；
Step 105: the implementing result of multiple sub- converging operations being merged, the execution of the target polymerization operation is obtained
As a result.
Specifically, merging to the implementing result of multiple sub- converging operations, the execution of the target polymerization operation is obtained
As a result, may include: the merging for carrying out the amount of offsetting to the implementing result of sub- converging operation, the target polymerization operation is obtained
Implementing result, wherein the offset of every sub- converging operation polymerize behaviour according to each height before being located at current sub- converging operation
The length of the sliding window of work determines.
In upper example, the target polymerization operation of the parameter area by sliding window length beyond accelerator is split as multiple cunnings
Dynamic length of window is all located at the sub- converging operation in the parameter area, then, then by accelerator to multiple sub- converging operations
Accelerated, obtain the implementing result of multiple sub- converging operations, finally the implementing result of multiple sub- converging operations is merged,
Obtain the implementing result of target polymerization operation.It is not high to solve existing accelerator data fitness through the above way, and leads
It is poly- in processing sliding window to have reached promotion accelerator for the problem of converging operation excessive to sliding window length caused can not be handled
The parameter adaptability of closing operation, to improve the technical effect of its performance for handling kernel function.
Above-mentioned target polymerization operation can be following one: accumulation operations, maximizing, minimize, ask at repeated subtraction
Average value.Above-mentioned accelerator can be the dedicated accelerator of time series, correspondingly, the above method can be applied to time series point
Analysis, sliding window converging operation especially in time series, and be related to during hardware-accelerated sliding window converging operation, such as
What design solves any length of window sliding window converging operation.
The above method is illustrated below with reference to a specific embodiment, it should be noted, however, that the specific implementation
Example does not constitute an undue limitation on the present application merely to the application is better described.
In order to improve the adaptability for accelerating verification sliding window size in the dedicated accelerator of time series, can be realized to any
Length of window can be carried out sliding window converging operation and be not only restricted to accelerator when realizing to the buffer area that each acceleration core is reserved
Size.A kind of converging operation processing method is provided in this example, may include steps of:
S1: the parameter area of the sliding window length of the dedicated accelerator of time series is determined；
S2: sliding window converging operation of the sliding window length parameter other than the range is resolved into multiple sub- sliding windows
Mouth converging operation, so that the sliding window length of each word sliding window converging operation is in the parameter area；
S3: accelerate the converging operation of the sub- sliding window using the dedicated accelerator of time series, obtain sub- converging operation
Operation result；
S4: the result of sub- converging operation is merged, and obtains the result of original sliding window converging operation.
That is, providing a kind of method for solving any length of window sliding window converging operation can have by sub- polymerization
Effect improves the dedicated accelerator of time series in the parameter usability of processing sliding window converging operation, and does not need to existing
Accelerator is redesigned.
The above method is specifically described with reference to the accompanying drawings and detailed description:
Assuming that the dedicated accelerator of time series has an acceleration core ts_sum, for ask sequence and, and only maximum is supported to ask 8
The core of a data.Such as: sequence A, the sum that window size is 8 are asked with ts_sum (A, 8) expression.Assuming that current window size is
When 18, then can realize that above-mentioned window size is 18 by existing acceleration core ts_sum (A, 8) in the following way
Sequence and.
Specifically, the off-limits kernel function calculating of window is decomposed into daughter nucleus functional operation, the window of each daughter nucleus function
For mouth length all within the scope of accelerator is supported, mathematical expression is as follows:
Ts_sum (A, 8)=ts_sum (A, 8)+forward (ts_sum (A, 8), 8)+forward (ts_sum (A, 2),
10)
It is illustrated in figure 2 the length of window size for accelerating core ts_sum to support, accelerates core to support as seen from Figure 2 each
The size of data that length is 8 is accessed, a corresponding output is then generated.Fig. 3 is when application requirement length of window is 18, now
Some accelerators cannot be supported.It can be more than to accelerate core in the length of window of time series based on above-mentioned provided method
Window size when, do not change accelerate to verify it is existing itself under the premise of, corresponding operation can be completed.
Specifically, entire operation can be split as to 4 sub-operations, Fig. 4 is sub-operation 1, and ts_sum (A, 8) completes window
Size is 8 operation, and generates sub- output sequence 1；Fig. 5 is sub-operation 2, and ts_sum (A, 8) completes other sequence A from the 9th
The window size that data start is 8 sliding window converging operation, and generates sub- output sequence 2；Fig. 6 is sub-operation 3, ts_sum
Window size of (A, 2) the deadline sequence A since the 17th data is 2 sliding window converging operation, and generates sub- output
Sequence 3.Fig. 7 is that sub- output sequence merges, wherein the output sequence union operation is the son output sequence for generating front sub-operation
Column add up, and carry out the merging for the amount of offsetting.As shown, when realizing, length of window n is equal to 8, therefore needs
By sub- output sequence 2 move forward 8 clock cycle and and by sub- output sequence 3 move forward 10 clock cycle, then by three sons
Output sequence is overlapped operation, to obtain result output sequence to the end.
It should be noted, however, that above-mentioned implementation method is not only adapted to this kind of set operation of above-mentioned ts_sum, it is all poly-
Closing operation meets invertibity, that is, exist operation+and-, so as to x and y in entire sequence, there is (x+y)-y=x establishment, such as:
Ask the sliding window converging operation a kind of in this way such as the maximum value of sequence designated length or ts_max, ts_min of minimum value can be direct
Use the method in this.
Certainly, reversible requirement may be unsatisfactory for there are also some converging operations.Such as: ask the sequence of designated length sequence
The ts_rank converging operation of value, although can not be calculated with reference to sequence using the calculation of the expense very little in this example
The fractionation of method " dichotomy " Lai Shixian sliding window converging operation.
By the method and apparatus of any length of window sliding window converging operation of realization of upper example, time series can be promoted
Dedicated accelerator is handling the parameter usability of sliding window converging operation, to improve its performance for handling kernel function.At this
In specification, such as adjective as first and second be can be only used for an element or movement and another element or movement
It distinguishes, without requiring or implying any actual this relationship or sequence.In the case where environment allows, referring to element
Or component or step (s) should not be interpreted as limited to one in only element, component or step, and can be element, component,
Or one or more in step etc..
According to embodiments of the present invention, a kind of target polymerization operation processing method embodiment is additionally provided, it should be noted that
Step shown in the flowchart of the accompanying drawings can execute in a computer system such as a set of computer executable instructions, and
It, in some cases, can be to be different from sequence execution institute herein and although logical order is shown in flow charts
The step of showing or describing.
Embodiment of the method provided by above-mentioned can be held in mobile terminal, terminal or similar arithmetic unit
Row.For running on computer terminals, Fig. 8 is a kind of calculating of target polymerization operation processing method of the embodiment of the present invention
The hardware block diagram of machine terminal.As shown in figure 8, terminal 10 may include one or more (only showing one in figure)
(processor 102 can include but is not limited to the processing dress of Micro-processor MCV or programmable logic device FPGA etc. to processor 102
Set), memory 104 for storing data and the transmission module 106 for communication function.Those of ordinary skill in the art
It is appreciated that structure shown in Fig. 8 is only to illustrate, the structure of above-mentioned electronic device is not caused to limit.For example, computer
Terminal 10 may also include than shown in Fig. 8 more perhaps less component or with the configuration different from shown in Fig. 8.
Memory 104 can be used for storing the software program and module of application software, such as the target in the embodiment of the present invention
Corresponding program instruction/the module of converging operation processing method, the software that processor 102 is stored in memory 104 by operation
Program and module realize the target polymerization of above-mentioned application program thereby executing various function application and data processing
Operation processing method.Memory 104 may include high speed random access memory, may also include nonvolatile memory, such as one or
Multiple magnetic storage devices, flash memory or other non-volatile solid state memories.In some instances, memory 104 can be into one
Step includes the memory remotely located relative to processor 102, these remote memories can pass through network connection to computer
Terminal 10.The example of above-mentioned network includes but is not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.
Transmission module 106 is used to that data to be received or sent via a network.Above-mentioned network specific example may include
The wireless network that the communication providers of terminal 10 provide.In an example, transmission module 106 includes that a network is suitable
Orchestration (Network Interface Controller, NIC), can be connected by base station with other network equipments so as to
Internet is communicated.In an example, transmission module 106 can be radio frequency (Radio Frequency, RF) module,
For wirelessly being communicated with internet.
Under above-mentioned running environment, in software view, target polymerization operation processing device can be as shown in Figure 9, comprising: obtains
Modulus block 901, splits module 903, accelerating module 904 and merging module 905 at determining module 902, in which:
Module 901 is obtained, the parameter area of the sliding window length for obtaining target polymerization operation and accelerator；
Determining module 902, for determining whether the sliding window length of the target polymerization operation exceeds the parameter model
It encloses；
Module 903 is split, for the target polymerization being operated and is split in the case where determining beyond the parameter area
For multiple sub- converging operations, wherein the sliding window length of every sub- converging operation is all located in the parameter area；
Accelerating module 904 obtains multiple sub- polymerizations for accelerating by accelerator to the multiple sub- converging operation
The implementing result of operation；
Merging module 905 is merged for the implementing result to multiple sub- converging operations, obtains the target polymerization behaviour
The implementing result of work.
In one embodiment, splitting module 903 may include: acquiring unit, for obtaining in the parameter area
Upper limit value；Split cells, for the target polymerization operation order to be split as N number of sub- converging operation, wherein preceding N-1
The length of the sliding window of sub- converging operation is the upper limit value, and the length of the sliding window of the sub- converging operation of n-th is less than etc.
In the upper limit value.
In one embodiment, merging module 905 specifically can be used for the implementing result of the sub- converging operation into
The merging of the row amount of offsetting obtains the implementing result of the target polymerization operation, wherein the offset root of every sub- converging operation
It is determined according to the length of the sliding window of each sub- converging operation before being located at current sub- converging operation.
In one embodiment, target polymerization operation can be following one: accumulation operations, repeated subtraction ask maximum
It is worth, minimize, averages.
In one embodiment, above-mentioned accelerator can be the dedicated accelerator of time series.
In another embodiment, a kind of software is additionally provided, the software is for executing above-described embodiment and preferred reality
Apply technical solution described in mode.
In another embodiment, a kind of storage medium is additionally provided, above-mentioned software is stored in the storage medium, it should
Storage medium includes but is not limited to: CD, floppy disk, hard disk, scratch pad memory etc..
It can be seen from the above description that the embodiment of the present invention realizes following technical effect: by sliding window length
The target polymerization operation of parameter area beyond accelerator is split as multiple sliding window length and is all located in the parameter area
Sub- converging operation, then, then by accelerator multiple sub- converging operations are accelerated, obtain holding for multiple sub- converging operations
Row obtains the implementing result of target polymerization operation as a result, finally merge to the implementing result of multiple sub- converging operations.Pass through
Aforesaid way solves that existing accelerator data fitness is not high, caused by the converging operation excessive to sliding window length
The problem of can not handling, has reached and has promoted accelerator in the parameter adaptability of processing sliding window converging operation, to improve it
Handle the technical effect of the performance of kernel function.
Although this application provides the method operating procedure as described in embodiment or flow chart, based on conventional or noninvasive
The labour for the property made may include more or less operating procedure.The step of enumerating in embodiment sequence is only numerous steps
One of execution sequence mode, does not represent and unique executes sequence.It, can when device or client production in practice executes
To execute or parallel execute (such as at parallel processor or multithreading according to embodiment or method shown in the drawings sequence
The environment of reason).
The device or module that above-described embodiment illustrates can specifically realize by computer chip or entity, or by having
The product of certain function is realized.For convenience of description, it is divided into various modules when description apparatus above with function to describe respectively.
The function of each module can be realized in the same or multiple software and or hardware when implementing the application.It is of course also possible to
Realization the module for realizing certain function is combined by multiple submodule or subelement.
Method, apparatus or module described herein can realize that controller is pressed in a manner of computer readable program code
Any mode appropriate is realized, for example, controller can take such as microprocessor or processor and storage can be by (micro-)
The computer-readable medium of computer readable program code (such as software or firmware) that processor executes, logic gate, switch, specially
With integrated circuit (Application Specific Integrated Circuit, ASIC), programmable logic controller (PLC) and embedding
Enter the form of microcontroller, the example of controller includes but is not limited to following microcontroller: ARC625D, Atmel AT91SAM,
Microchip PIC18F26K20 and Silicone Labs C8051F320, Memory Controller are also implemented as depositing
A part of the control logic of reservoir.It is also known in the art that in addition to real in a manner of pure computer readable program code
Other than existing controller, completely can by by method and step carry out programming in logic come so that controller with logic gate, switch, dedicated
The form of integrated circuit, programmable logic controller (PLC) and insertion microcontroller etc. realizes identical function.Therefore this controller
It is considered a kind of hardware component, and hardware can also be considered as to the device for realizing various functions that its inside includes
Structure in component.Or even, it can will be considered as the software either implementation method for realizing the device of various functions
Module can be the structure in hardware component again.
Part of module in herein described device can be in the general of computer executable instructions
Upper and lower described in the text, such as program module.Generally, program module includes executing particular task or realization specific abstract data class
The routine of type, programs, objects, component, data structure, class etc..The application can also be practiced in a distributed computing environment,
In these distributed computing environment, by executing task by the connected remote processing devices of communication network.In distribution
It calculates in environment, program module can be located in the local and remote computer storage media including storage equipment.
As seen through the above description of the embodiments, those skilled in the art can be understood that the application can
It is realized by the mode of software plus required hardware.Based on this understanding, the technical solution of the application is substantially in other words
The part that contributes to existing technology can be embodied in the form of software products, and can also pass through the implementation of Data Migration
It embodies in the process.The computer software product can store in storage medium, such as ROM/RAM, magnetic disk, CD, packet
Some instructions are included to use so that a computer equipment (can be personal computer, mobile terminal, server or network are set
It is standby etc.) execute method described in certain parts of each embodiment of the application or embodiment.
Each embodiment in this specification is described in a progressive manner, the same or similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.The whole of the application or
Person part can be used in numerous general or special purpose computing system environments or configuration.Such as: personal computer, server calculate
Machine, handheld device or portable device, mobile communication terminal, multicomputer system, based on microprocessor are at laptop device
System, programmable electronic equipment, network PC, minicomputer, mainframe computer, the distribution including any of the above system or equipment
Formula calculates environment etc..
Although depicting the application by embodiment, it will be appreciated by the skilled addressee that the application there are many deformation and
Variation is without departing from spirit herein, it is desirable to which the attached claims include these deformations and change without departing from the application's
Spirit.
Claims (9)
1. a kind of target polymerization operation processing method, which is characterized in that method includes the following steps:
Obtain the parameter area of the sliding window length of target polymerization operation and accelerator；
Determine whether the sliding window length of the target polymerization operation exceeds the parameter area；
In the case where determining beyond the parameter area, target polymerization operation is split as multiple sub- converging operations,
In, the sliding window length of every sub- converging operation is all located in the parameter area；
The multiple sub- converging operation is accelerated by accelerator, obtains the implementing result of multiple sub- converging operations；
The implementing result of multiple sub- converging operations is merged, the implementing result of the target polymerization operation is obtained.
2. being grasped the method according to claim 1, wherein target polymerization operation is split as multiple sub- polymerizations
As the step of include:
Obtain the upper limit value in the parameter area；
The target polymerization operation order is split as N number of sub- converging operation, wherein the sliding window of preceding N-1 sub- converging operations
The length of mouth is the upper limit value, and the length of the sliding window of the sub- converging operation of n-th is less than or equal to the upper limit value.
3. the method according to claim 1, wherein merged to the implementing result of multiple sub- converging operations,
The step of obtaining the implementing result of target polymerization operation include:
The merging that the amount of offsetting is carried out to the implementing result of the sub- converging operation obtains the execution knot of the target polymerization operation
Fruit, wherein the offset of every sub- converging operation is according to the cunning of each sub- converging operation before being located at currently sub- converging operation
The length of dynamic window determines.
4. the method according to claim 1, wherein target polymerization operation is one of the following: accumulation operations,
Repeated subtraction maximizing, minimizes, averages.
5. method according to claim 1 to 4, which is characterized in that the accelerator is that time series is dedicated
Accelerator.
6. a kind of target polymerization operation processing device, which is characterized in that the device includes:
Module is obtained, the parameter area of the sliding window length for obtaining target polymerization operation and accelerator；
Determining module, for determining whether the sliding window length of the target polymerization operation exceeds the parameter area；
Module is split, for target polymerization operation being split as multiple in the case where determining beyond the parameter area
Sub- converging operation, wherein the sliding window length of every sub- converging operation is all located in the parameter area；
Accelerating module obtains multiple sub- converging operations for accelerating by accelerator to the multiple sub- converging operation
Implementing result；
Merging module is merged for the implementing result to multiple sub- converging operations, obtains holding for the target polymerization operation
Row result.
7. device according to claim 6, which is characterized in that the fractionation module includes:
Acquiring unit, for obtaining the upper limit value in the parameter area；
Split cells, for the target polymerization operation order to be split as N number of sub- converging operation, wherein preceding N-1 son polymerization
The length of the sliding window of operation is the upper limit value, and the length of the sliding window of the sub- converging operation of n-th is less than or equal to described
Upper limit value.
8. device according to claim 6, which is characterized in that the merging module is specifically used for the sub- converging operation
Implementing result carry out the merging of the amount of offsetting, obtain the implementing result of the target polymerization operation, wherein every height polymerization behaviour
The offset of work is determined according to the length of the sliding window of each sub- converging operation before being located at current sub- converging operation.
9. a kind of non-volatile computer readable storage medium storing program for executing, is stored thereon with computer program, which is characterized in that the calculating
Machine program realizes the step of any one of claims 1 to 5 the method when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811158944.6A CN109408545B (en) | 2018-09-30 | 2018-09-30 | Target polymerization operation processing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811158944.6A CN109408545B (en) | 2018-09-30 | 2018-09-30 | Target polymerization operation processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109408545A CN109408545A (en) | 2019-03-01 |
| CN109408545B true CN109408545B (en) | 2019-10-25 |
Family
ID=65465952
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811158944.6A Active CN109408545B (en) | 2018-09-30 | 2018-09-30 | Target polymerization operation processing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109408545B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102682086A (en) * | 2012-04-23 | 2012-09-19 | 华为技术有限公司 | Data segmentation method and data segmentation equipment |
| CN103309873A (en) * | 2012-03-09 | 2013-09-18 | 阿里巴巴集团控股有限公司 | Method and device for processing data, and system |
| US8548900B1 (en) * | 2012-12-19 | 2013-10-01 | Nyse Group, Inc. | FPGA memory paging |
| CN104809333A (en) * | 2015-04-03 | 2015-07-29 | 百度在线网络技术（北京）有限公司 | Capacity predicating method and system based on Kalman filter |
| CN107274009A (en) * | 2017-05-27 | 2017-10-20 | 中国科学院计算技术研究所 | A kind of time series data multistep forecasting method and system based on correlation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106960059A (en) * | 2017-04-06 | 2017-07-18 | 山东大学 | A kind of Model of Time Series Streaming dimensionality reduction based on Piecewise Linear Representation is with simplifying method for expressing |
-
2018
- 2018-09-30 CN CN201811158944.6A patent/CN109408545B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103309873A (en) * | 2012-03-09 | 2013-09-18 | 阿里巴巴集团控股有限公司 | Method and device for processing data, and system |
| CN102682086A (en) * | 2012-04-23 | 2012-09-19 | 华为技术有限公司 | Data segmentation method and data segmentation equipment |
| US8548900B1 (en) * | 2012-12-19 | 2013-10-01 | Nyse Group, Inc. | FPGA memory paging |
| CN104809333A (en) * | 2015-04-03 | 2015-07-29 | 百度在线网络技术（北京）有限公司 | Capacity predicating method and system based on Kalman filter |
| CN107274009A (en) * | 2017-05-27 | 2017-10-20 | 中国科学院计算技术研究所 | A kind of time series data multistep forecasting method and system based on correlation |
Non-Patent Citations (1)
| Title |
|---|
| 基于并行计算的数据流处理方法研究;周勇;《中国博士学位论文全文数据库 信息科技辑》;20140515;第I138-29页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109408545A (en) | 2019-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3267310A1 (en) | Data processing method and device | |
| CN107678865A (en) | The verification method and system of block chain based on transaction packet | |
| CN107196900A (en) | A kind of method and device for verification of knowing together | |
| US11501160B2 (en) | Cloud computing data compression for allreduce in deep learning | |
| US8595735B2 (en) | Holistic task scheduling for distributed computing | |
| JP2022511613A (en) | Co-scheduling of quantum computing jobs | |
| CN110086650B (en) | Cloud resource online scheduling method and device for distributed machine learning task | |
| CN107729137B (en) | Server, block chain signature verification decryption method and storage medium | |
| CN111198754B (en) | Task scheduling method and device | |
| CN109117141A (en) | Simplify method, apparatus, the electronic equipment, computer readable storage medium of programming | |
| CN107783731B (en) | Big data real-time processing method and system | |
| CN106373112A (en) | Image processing method, image processing device and electronic equipment | |
| CN107798039A (en) | A kind of method of data synchronization and device | |
| US9529572B1 (en) | Composable application session parameters | |
| CN109669773A (en) | Finance data processing method, device, equipment and storage medium | |
| Xu et al. | Theoretical results of QoS-guaranteed resource scaling for cloud-based MapReduce | |
| CN109408545B (en) | Target polymerization operation processing method and device | |
| CN110442647B (en) | Data consistency synchronization method and device and computer readable storage medium | |
| CN110162398A (en) | A kind of dispatching method, device and the terminal device of diseases analysis model | |
| CN109508412B (en) | Method and device for constructing computation flow graph processed by time series | |
| Kim et al. | Towards the design of a system and a workflow model for medical big data processing in the hybrid cloud | |
| CN109388609A (en) | Based on the data processing method and device for accelerating core | |
| CN115712422A (en) | Form page generation method and device, computer equipment and storage medium | |
| US11294704B2 (en) | Monitoring and reporting performance of online services using a monitoring service native to the online service | |
| CN112783949A (en) | Human body data prediction method and device, electronic equipment and storage medium |
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 |