CN102546203B - Business process allocation method and device - Google Patents
Business process allocation method and device Download PDFInfo
- Publication number
- CN102546203B CN102546203B CN201010595444.6A CN201010595444A CN102546203B CN 102546203 B CN102546203 B CN 102546203B CN 201010595444 A CN201010595444 A CN 201010595444A CN 102546203 B CN102546203 B CN 102546203B
- Authority
- CN
- China
- Prior art keywords
- service request
- business process
- time interval
- business
- user
- 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
Images
Abstract
The invention discloses a business process allocation method and a business process allocation device, which are used for solving the problems that business report generation is delayed and resources are wasted in the business process due to allocation of the business process. The business process allocation method includes: forecasting the business request quantity sent from users within the next time interval according to the business request quantity sent by users within each set time interval, and then regulating the request quantity of business process for processing the business the next time according to the forecast business request quantity and the received business request quantity sent by users at present time. According to the technical scheme, the business request quantity sent by users possibly the next time is forecast, and the request quantity of business process for processing the business is regulated according to the forecasting results dynamically, so that the requirement for time limit during generation of business reports is guaranteed, the generated business reports are more accurate, and waste of resources in the business process can be avoided.
Description
Technical field
The present invention relates to business support technical field, relate in particular to a kind of business process distribution method and device.
Background technology
Business support system provides the statistical functions such as the situation of distributing telephone numbers, payment situation and arrearage situation, and user, can be according to the service request that need to send of self in order to grasp production and operation situation, asks corresponding statistical report form.Due to generating when statistical report form, the concrete condition of business support system now need to be informed to user, therefore report generation has higher time limit requirement.In order to meet the time limit requirement of report generation, different service request has been distributed to corresponding business process, every kind of service request is processed by corresponding business process.
Be all static for the process of service request distribution service process in the prior art, distribute the business process of fixed qty for certain class service request.But the transmission of user's service request has larger time bursts, there is larger gap in the quantity of the service request sending business busy and business idle user.Therefore when having distributed after the business process of respective numbers for a certain service request, the service request sending as user is when more, the business process of this fixed qty may be able to not be processed the service request of this quantity, thereby cause traffic backlog, the bordereau generating can accurately not react current state of affairs, the service request sending as user when less, can cause the residue of business process, thereby cause the wasting of resources.
At present, in order to address the above problem, also be the quantity that can increase dynamically or reduce business process, but this process that dynamically increases or reduce business process is passive carrying out, in the time detecting that current business process can not meet the needs of bordereau generation, increase the quantity of business process.But the quantity that now increases again business process has caused report generation to have time delay, can accurately not react current state of affairs.
Therefore in prior art, owing to not distributing corresponding business process according to the quantity of service request, cause the generation of bordereau to have time delay, can accurately not react current service condition, in addition, also can distribute too much because of business process, cause the wasting of resources of business process.
Summary of the invention
In view of this, the embodiment of the present invention provides a kind of business process distribution method and device, cause the generation of bordereau to have time delay in order to solve the distribution of prior art business process, can accurately not react current service condition, and cause the problem of the wasting of resources of business process.
A kind of business process distribution method that the embodiment of the present invention provides, comprising:
The service request quantity sending according to the user who has received in the time interval of each setting, the service request quantity that in next time interval of prediction current time interval, user sends;
In the time that the described service request quantity of judgement prediction is greater than the service request quantity that in current time interval, user sends, increase the quantity allotted of business process;
In the time that the described service request quantity of judgement prediction is less than the service request quantity that in current time interval, user sends, reduce the quantity allotted of business process.
A kind of business process distributor that the embodiment of the present invention provides, comprising:
Prediction module, for the service request quantity sending according to the user who has received in the time interval of each setting, the service request quantity that in next time interval of prediction current time interval, user sends;
Distribution module, for in the time that the described service request quantity of judgement prediction is greater than the service request quantity that in current time interval, user sends, increase the quantity allotted of business process, in the time that the described service request quantity of judgement prediction is less than the service request quantity that in current time interval, user sends, reduce the quantity allotted of business process.
The embodiment of the present invention provides a kind of business process distribution method and device, the service request quantity that the method sends according to the user who has received in the time interval of each setting, the service request quantity that user in next time interval of current time interval is sent is predicted, according to the service request quantity predicting, and the quantity of the service request that sends of the user that receives of current time, adjust the quantity of the business process of in next time interval, this service request being processed.Due to the service request quantity having sent according to user in embodiments of the present invention, the service request quantity that in next time interval of current time interval, user may send is predicted, and according to the quantity of the business process that dynamically adjustment is processed service request that predicts the outcome, thereby can ensure the time limit requirement that bordereau generates, make the bordereau of generation more accurate, in addition due to according to the quantity of service request, business process is distributed, also can not cause the wasting of resources of business process.
Brief description of the drawings
Service request prediction and business process assigning process that Fig. 1 provides for the embodiment of the present invention;
The specific implementation process that service request is predicted and business process distributes that Fig. 2 provides for the embodiment of the present invention;
The structural representation of the business process distributor that Fig. 3 provides for the embodiment of the present invention.
Embodiment
The embodiment of the present invention is in order effectively to reduce the time delay of report generation, improve the accuracy that form improves, a kind of service request prediction and business process distribution method and device are provided, the quantity of the service request that the user that the method has received according to each setting-up time interval sends, the service request quantity that user in next time interval is sent is predicted, according to the service request quantity predicting, and the quantity of the service request that sends of the user that receives of current time, adjust the quantity of the business process of in next time interval, this service request being processed.Due to the service request quantity having sent according to user in embodiments of the present invention, the service request quantity that may send user in next time interval is predicted, and according to the quantity of the business process that dynamically adjustment is processed service request that predicts the outcome, thereby can ensure the time limit requirement that bordereau generates, make the bordereau of generation more accurate, due to according to the quantity of service request, business process is distributed in addition, also can not cause the wasting of resources of business process.
Below in conjunction with Figure of description, the embodiment of the present invention is elaborated.
Service request prediction and business process assigning process that Fig. 1 provides for the embodiment of the present invention, this process comprises the following steps:
S101: the service request quantity sending according to user in the time interval of each setting, the service request quantity that in next time interval of prediction current time interval, user sends.
The service request quantity that in next time interval of concrete prediction current time interval, user sends comprises:
The service request quantity that adopts the interior user of next time interval of difference autoregression rolling average (Autoregressive Integrated Moving Average Model, ARIMA) model prediction current time interval to send, comprises and adopts Y
t=α
0+ α
1y
t-1+ ... + α
py
t-p+ β
0μ
t+ β
1μ
t-1+ ... + β
qμ
t-qpredict the service request quantity that in next time interval, user sends, wherein Y
tfor the service request quantity that in next time interval of prediction, user sends, Y
t-1y
t-pfor the service request quantity that user in this each this setting-up time interval sends, α
0α
pfor autoregressive coefficient, β
0β
qfor moving average coefficient, μ
tμ
t-qfor rolling average item, the son that p is model returns item number, the rolling average item number that q is model.
S102: whether the described service request quantity that judges prediction is greater than the service request quantity that user sends in current time interval, when judged result is when being, carries out step S103, otherwise, carry out step S104.
S103: the quantity allotted that increases the business process that service request is processed.
S104: the quantity allotted that reduces the business process that service request is processed.
In embodiments of the present invention the service request quantity sending in next time interval of user's current time interval is predicted, and according to the quantity of the result adjustment business process that service request is processed of prediction, thereby reach the object of dynamic assignment business process.And the quantity of service request that user sends within to next time interval is while predicting, can adopt other model, or the trend of the service request quantity sending according to the user who has received in the time interval of each setting, the quantity of business process is adjusted.The service request quantity for example sending as the user who has received in the time interval of this each setting is what increase progressively, predict the quantity of the service request that in next time interval, user sends, also increase to some extent compared with current time interval, the quantity increasing, can select arbitrarily, or the service request quantity sending according to the user who has received in this each time interval is predicted.Certainly the accuracy in order to predict, can adopt ARIMA model.
The service request quantity that in concrete each setting-up time interval, user sends can be considered as a random sequence, this time interval can be set as required flexibly, can be for example 5 minutes, 10 minutes etc., and can prove that the service request quantity that this each time interval user sends meets ARIMA model requirements.Therefore the service request quantity that adopts ARIMA model to send user in next time interval is predicted.
This ARIMA model can use ARIMA (p, d, q) to represent, wherein AR refers to autoregression, the autoregression item number that p is model, and MA is rolling average, q is the rolling average item number of model, and I refers to integration, and d is the number of times that must get its difference between time series becomes steadily.This ARIMA model can represent with following formula:
Y
t＝α
0+α
1Y
t-1+……+α
pY
t-p+β
0μ
t+β
1μ
t-1+……+β
qμ
t-q
Wherein Y
tfor the service request quantity that in next time interval of prediction, user sends, Y
t-1y
t-pfor the service request quantity that user in each this setting-up time interval sends, α
0α
pfor autoregressive coefficient, β
0β
qfor moving average coefficient, μ
tμ
t-qfor rolling average item, the son that p is model returns item number, the rolling average item number that q is model.
Describe the schedule of quantities of the service request that table 1 sends for the user who receives in 8:00～8:20 time interval of every 5 minutes below with a specific embodiment.The quantity of the service request that need to send to user in this time interval of 8:25 at 8:20 user is now predicted.
Table 1
Adopt ARIMA model to user in the time that 8:20 predicts to the quantity of the service request sending in this time interval of 8:25, the quantity of the service request of the user's transmission receiving within each time interval of statistics in this table 1 is as input parameter.ARIMA model, according to the information of the parameter of input, carries out data derivation prediction.
First need to carry out balance processing, this is because the quantity of the service request that the user who receives in these time intervals of 8:00-8:20 sends is sustainable growth, and amplification is obvious.ARIMA model can be taken from right logarithm operation to the parameter of input, to data Y
t-iget natural logrithm, make X
t-i=ln Y
t-i, and to getting the X after natural logrithm
tcarry out first difference computing Z
t=X
t-X
t-1, one group of data after calculus of differences meet ARIMA model and carry out subsequent calculations condition.
In said process, ARIMA model has carried out first difference processing to the data of input, therefore between becoming steadily, must to get the number of times d of its difference be 1 to time series, when definite auto-correlation item number p=4, the rolling average item number q=4 of ARIMA model, the model image of matching and the data fitting of 8:00-8:20 obtain best, therefore can determine Z
t=0.1832+0.5229Z
t-4-0.945 μ
t-4, because Z
t=X
t-X
t-1, X
t-i=ln Y
t-i, therefore can determine Y
t=Y
t-1+ Z
t=Y
t-1+ 0.1832+0.5229Z
t-4-0.945 μ
t-4, the result that the quantity of the service request therefore user that may receive in next time interval 8:20～8:25 being sent is predicted is 1482.
And adopt above-mentioned ARIMA model can predict the quantity of the service request of the user's transmission receiving in the each time interval in 8:00～8:20 too, as shown in table 2 below:
| Time | 8:00 | 8:05 | 8:10 | 8:15 | 8:20 | 8:25 |
| Actual value | 622 | 815 | 882 | 995 | 1202 | Nothing temporarily |
| Predicted value | 632 | 756 | 869 | 1011 | 1195 | 1482 |
Table 2
Can see from above-mentioned table 2, the quantity of the service request that the actual user who receives sends within each time interval, the variance rate of directly pressing with the quantitative forecast value of the service request to user may send in this time interval, within 9%, therefore can prove that the quantity of the service request that this model can send user in next time interval of current time interval accurately be predicted.
When predicting after the service request quantity that in next time interval, user sends, in order effectively the service request of this quantity to be processed, meet the time limit requirement of report generation, and do not cause the waste of resource yet, need in embodiments of the present invention according to the quantity of the service request that in next time interval predicting, user sends, and the service request quantity that in current time interval, user sends, the quantity of processing the business process that service request that user in this next time interval sends processes is adjusted.
In the time that the quantity of the business process that the service request of processing user's transmission in this next time interval is processed is adjusted, first need the quantity of the service request that judges that in next time interval of predicting, user may send whether to be greater than the service request quantity that in current time interval, user sends, in the time that the described service request quantity of judgement prediction is greater than the service request quantity that in current time interval, user sends, need to increase the quantity allotted of business process, in the time that the described service request quantity of judgement prediction is less than the service request quantity that in current time interval, user sends, reduce the quantity allotted of business process.
Or in the time judging, can preserve the mean value of the service request quantity that in the time interval, user sends, and the quantity of the business process that the mean value of this service request quantity is processed, according to result relatively, the service request quantity that next time interval distributes is adjusted.
In the time that the described service request quantity of judgement prediction is greater than the service request quantity of current time user transmission, the quantity that increases business process comprises: the business who determines service request quantity and the service request quantity that current time user sends of prediction, according to amassing of this business who determines and current number of processes of having distributed, determine the quantity of the business process increasing; And according to definite quantity, increase the quantity allotted of business process.
Before increasing the quantity of business process, described method also comprises: the quantity that judges the current business process having distributed is less than the threshold value that business process distributes.
Wherein, after determining the quantity of the business process increasing, also comprise: whether the quantity that judges the business process after increasing is greater than the threshold value that business process distributes; When the quantity of business process after increasing is greater than the threshold value that business process distributes, the threshold value of distributing according to business process and current number of processes of having distributed poor, redefines the quantity of the business process of increase.
When business process quantity after increasing is not more than the threshold value of business process, increase the quantity of this definite business process.
In the time that the described service request quantity of judgement prediction is less than the service request quantity of current time user transmission, the quantity allotted that reduces business process comprises: determine the business of service request quantity and the service request quantity that current time user sends of prediction, and definite predefined parameter and described business's is poor; According to amassing of described difference and current number of processes of having distributed, determine the quantity of the business process reducing; According to definite quantity, reduce the quantity allotted of business process.
The specific implementation process that service request is predicted and business process distributes that Fig. 2 provides for the embodiment of the present invention, this process comprises the following steps:
S201: the service request quantity sending according to user in the time interval of each setting, adopts difference ARMA model to predict the service request quantity that in next time interval, user sends
S202: whether the described service request quantity that judges prediction is greater than the service request quantity that current time user sends, and when judged result is when being, carries out step S203, otherwise, carry out step S208.
S203: judge whether the business process quantity of current distribution is not more than the threshold value of business process, when judged result is when being, carry out step S204, otherwise, carry out step S209.
S204: determine the business of service request quantity and the service request quantity that current time user sends of prediction, according to amassing of the number of processes of this business who determines and current distribution, determine the quantity of the business process increasing.
S205: judge whether the quantity of business process after increasing is greater than the threshold value of business process, when judged result is when being, carry out step S206, otherwise, carry out step S207.
S206: poor according to the number of processes of the threshold value of business process and current distribution, determine the quantity of the business process increasing.
S207: the quantity allotted that increases this definite business process.
S208: determine the business of the service request quantity that service request quantity and the current time user of prediction send, and definite predefined parameter and described business's is poor, according to amassing of the number of processes of described difference and current distribution, the quantity of the business process of definite minimizing.
S209: keep the quantity of business process of this distribution constant, wait for after the time interval of setting, carry out step S201.
In embodiments of the present invention, when having predicted after the quantity of the service request that in next time interval, user sends, and judge when service request quantity that next time interval user may send is less than the service request quantity that the user that receives in current time interval sends, the quantity of the business process that need to reduce service request to process, the concrete quantity reducing is N, the quantity of the process of N=(the service request quantity that the user that service request quantity/current time interval that next time interval user of 1-prediction sends receives sends) × current distribution.
In the time judging that service request quantity that next time interval user may send is greater than the service request quantity that the user that receives in current time interval sends, need to increase the quantity of the business process that service request is processed, at that time before increasing the quantity of business process, first need whether be not more than this system in the quantity of the business process that judges current distribution and allow the threshold value of the business process distributing, in the time judging that the quantity of business process of current distribution has equaled this system and allows the threshold value of the business process distributing, the quantity of the business process of its distribution can not be increased again, can only keep the business process quantity of distribution constant, after waiting for the time interval of setting, the quantity of the service request that the user in next time interval is sent is predicted, in the time judging that the quantity of business process of current distribution is less than this system and allows the threshold value of the business process distributing, increase the business process of respective numbers, the quantity of the business process of concrete increase is M, the quantity of the process of M=(the service request quantity that the user that service request quantity/current time interval that next time interval user of prediction sends receives sends) × current distribution.
In addition, limited because system allows the business process quantity of distributing, the quantity of the business process after increase can not be greater than the threshold value of the business process of this system permission distribution, therefore need to increase after the business process of respective numbers M when having determined, need to judge whether the quantity of the business process after increasing is greater than the threshold value of the business process of system permission distribution, when the quantity of the business process after increasing is greater than the threshold value of business process of system permission distribution, the quantity of this business process can only be increased to the threshold value of this business process, the quantity of the business process therefore now increasing is M-T, T is the threshold value of business process.When the quantity of the business process after increasing is not more than the threshold value of business process of system permission distribution, determine that the quantity of the business process increasing is M.
When the quantity of the service request that may send according to the user in next time interval of prediction, after business process is adjusted, there is the situation that increases or reduce business process, in the time that needs increase new business process, need to be this newly-increased business process distribution service request, or the service request of distributing to the business process having reclaimed is assigned on other business process.And every kind of service request of each business process processing exists very big-difference, also there is a great difference to the duration of every kind of service request processing in each business process therefore.
In the process of distributing in service request, embody fully the advantage of every kind of service request of each business process processing, improve the treatment effeciency of service request, need to add up the average time of every kind of service request of each business process processing; For every kind of service request, described service request is assigned to and is processed in the shortest business process described service request average time.
In embodiments of the present invention in order to embody the demand of load balancing and high-throughput two aspects, dividing timing to service request, need to consider the average time of every kind of service request of each business process processing, when time that a certain business process is processed a certain service request very in short-term, this kind of service request can be assigned in this business process and process, by this service request priority allocation to minimum business process of processing time, to ensure the load balancing of service request processing.
When every kind of service request has been distributed to after each business process, in order effectively and timely each service request to be processed, for every kind of service request, the business process being assigned to according to this service request is processed the time of this service request at every turn, and this business process processes the average time of this service request, determine this service request priority of this business process relatively; According to the priority of every kind of service request, adopt this business process to process each service request that it is assigned with.Each business process, according to the priority of each service request of determining, is processed each service request.
Concrete in embodiments of the present invention in order effectively to embody the principle of load balancing, and the treatment principle of high-throughput, priority minimum (Priority Min-Min, PMM) deadline algorithm the earliest can be adopted.Based on this PMM algorithm, need to determine that each business process processes the average handling time of every kind of service request, in the time determining that each business process is processed the average handling time of every kind of service request, can be according to following formula:
determine, wherein e
ikfor each processing time that this business process is processed this service request, m is the quantity that this business process is processed this service request,
for the average time of this business process to this service request processing, for the ease of calculating, the time that can process recently this service request according to this business process calculates.
In the time having determined that this business process is processed the shortest time of a certain service request, the service request of type in this all can be distributed to this business process and process.In the time that this business process is processed the service request of distributing to self, need to according to every kind of service request relatively the priority of this business process carry out.Concrete determining that every kind of service request is relatively when the priority of every kind of business process, the time that need at every turn process according to this business process this service request, and this business process is processed and average time of this service request is determined.Concrete can be according to formula:
determine every kind of service request priority of every kind of business process relatively, wherein Vi is the priority of definite relative i the business process of service request, e
ikfor each processing time that this business process is processed this service request, m is the quantity that this business process is processed this service request,
for the average time of this business process to this service request processing.
Describe with a specific embodiment below, the service request of for example carrying out business process distribution comprises that two kinds are respectively the service request of category-A type and the service request of category-B type, wherein, the service request of category-A type but before have 5 business process process, the service request of category-B type but before have 3 business process processing, wherein each time of processing each service request of each business process is respectively as shown in following table 3 and table 4:
| Number of times | Process 1 | Process 2 | Process 3 | Process 4 | Process 5 |
| For the first time | 100s | 121s | 154s | 87s | 112s |
| For the second time | 152s | 191s | 180s | 175s | 227s |
| For the third time | 189s | 302s | 208s | 145s | 351s |
The time of the service request processing of the each business process of table 3 to category-A type
| Number of times | Process 1 | Process 2 | Process 3 |
| For the first time | 220s | 321s | 254s |
| For the second time | 252s | 391s | 280s |
| For the third time | 389s | 352s | 408s |
The time of the service request processing of the each business process of table 4 to category-B type
Shown in above-mentioned table 3 and table 4, be respectively business process 1 the known average time that the service request of A type is processed for 147s, business process 2 is 204.7s, and business process 3 is 180.7s, and business process 4 is 135.7s, and business process 5 is 230.Be respectively business process 1 average time that the service request of B type is processed for 287s, business process 2 is 354.7s, and business process 3 is 314s.
Therefore known, business process 4 is the shortest to the average handling time of the service request of A type, therefore the service request of A type is assigned in business process 4, business process 1 is the shortest to the average handling time of the service request of B type, therefore the service request of B type is assigned in business process 1.And can calculate, the priority of the relative business process 4 of service request of A type is 6, the priority of the relative business process 1 of service request of B type is 8, therefore in the time that business process 4 and business process 1 are processed the service request of A type and B type, can process according to definite priority.
The structural representation of the business process distributor that Fig. 3 provides for the embodiment of the present invention, this device comprises:
Described prediction module 31 specifically for, adopt difference ARMA model to predict the service request quantity that in next time interval, user sends, comprise and adopt Y
t=α
0+ α
1y
t-1+ ... + α
py
t-p+ β
0μ
t+ β
1μ
t-1+ ... + β
qμ
t-qpredict the service request quantity that in next time interval, user sends, wherein Y
tfor the service request quantity that in next time interval of prediction, user sends, Y
t-1y
t-pfor the service request quantity that user in each this setting-up time interval sends, α
0α
pfor autoregressive coefficient, β
0β
qfor moving average coefficient, μ
tμ
t-qfor rolling average item, the son that p is model returns item number, the rolling average item number that q is model.
Described distribution module 32 specifically for, determine the business of service request quantity and the service request quantity that the interior user of current time interval sends of prediction, according to amassing of this business who determines and current number of processes of having distributed, determine the quantity of the business process increasing, and, according to definite quantity, increase the quantity allotted of business process.
Described distribution module 32 specifically for, whether the quantity that judges the business process after increasing is greater than the threshold value that business process distributes, when the quantity of the business process after increasing is greater than the threshold value of business process distribution, the threshold value of distributing according to business process and current number of processes of having distributed poor, redefine the quantity of the business process of increase.
Described distribution module 32 specifically for, determine the business of service request quantity and the service request quantity that the interior user of current time interval sends of prediction, and definite predefined parameter and described business's is poor, according to amassing of described difference and current number of processes of having distributed, determine the quantity of the business process reducing, according to definite quantity, reduce the quantity allotted of business process.
Described distribution module 32 is also for, average time of adding up each business process and processing every kind of service request, for every kind of service request, described service request is assigned to and processed in the shortest business process described service request average time.
Described distribution module 32 also for, for every kind of service request, the business process being assigned to according to this service request is processed the time of this service request at every turn, and this business process is processed the average time of this service request, determine this service request priority of this business process relatively, according to the priority of every kind of service request, adopt this business process to process each service request that it is assigned with.
The embodiment of the present invention provides a kind of business process distribution method and device, the service request quantity that the method sends according to the user who has received in the time interval of each setting, the service request quantity that user in next time interval of current time interval is sent is predicted, according to the service request quantity predicting, and the quantity of the service request that sends of the user that receives of current time, adjust the quantity of the business process of in next time interval, this service request being processed.Due to the service request quantity having sent according to user in embodiments of the present invention, the service request quantity that in next time interval of current time interval, user may send is predicted, and according to the quantity of the business process that dynamically adjustment is processed service request that predicts the outcome, thereby can ensure the time limit requirement that bordereau generates, make the bordereau of generation more accurate, in addition due to according to the quantity of service request, business process is distributed, also can not cause the wasting of resources of business process.
Obviously, those skilled in the art can carry out various changes and modification and not depart from the spirit and scope of the present invention the present invention.Like this, if these amendments of the present invention and within modification belongs to the scope of the claims in the present invention and equivalent technologies thereof, the present invention is also intended to comprise these changes and modification interior.
Claims (12)
1. a business process distribution method, is characterized in that, comprising:
The service request quantity sending according to the user who has received in the time interval of each setting, the service request quantity that in next time interval of prediction current time interval, user sends;
In the time that the described service request quantity of judgement prediction is greater than the service request quantity that in current time interval, user sends, determine the business of service request quantity and the service request quantity that the interior user of current time interval sends of prediction, according to amassing of this business who determines and current number of processes of having distributed, determine the quantity of the business process increasing, and according to definite quantity, increase the quantity allotted of business process;
In the time that the described service request quantity of judgement prediction is less than the service request quantity that in current time interval, user sends, determine the business of service request quantity and the service request quantity that the interior user of current time interval sends of prediction, and definite predefined parameter and described business's is poor, according to amassing of described difference and current number of processes of having distributed, determine the quantity of the business process reducing, according to definite quantity, reduce the quantity allotted of business process.
2. the method for claim 1, is characterized in that, the service request quantity that in next time interval of described prediction current time interval, user sends comprises:
Adopt difference ARMA model to predict the service request quantity that in next time interval, user sends, comprise and adopt Y
t=α
0+ α
1y
t-1+ ... + α
py
t-p+ β
0μ
t+ β
1μ
t-1+ ... + β
qμ
t-qpredict the service request quantity that in next time interval, user sends, wherein Y
tfor the service request quantity that in next time interval of prediction, user sends, Y
t-1y
t-pfor the service request quantity that user in each this setting-up time interval sends, α
0α
pfor autoregressive coefficient, β
0β
qfor moving average coefficient, μ
tμ
t-qfor rolling average item, the son that p is model returns item number, the rolling average item number that q is model.
3. the method for claim 1, is characterized in that, before the quantity of described increase business process, described method also comprises:
The quantity that judges the current business process having distributed is less than the threshold value that business process distributes.
4. the method for claim 1, is characterized in that, after the quantity of the business process of described definite increase, also comprises:
Whether the quantity that judges the business process after increasing is greater than the threshold value that business process distributes;
When the quantity of business process after increasing is greater than the threshold value that business process distributes, the threshold value of distributing according to business process and current number of processes of having distributed poor, redefines the quantity of the business process of increase.
5. the method as described in claim 1 or 4, is characterized in that, is increasing or is reducing after the quantity allotted of business process, and described method also comprises:
Add up each business process and process the average time of every kind of service request;
For every kind of service request, described service request is assigned to and is processed in the shortest business process described service request average time.
6. method as claimed in claim 5, is characterized in that, after described service request is assigned to and processes in the shortest business process described service request average time, described method also comprises:
For every kind of service request, the business process being assigned to according to this service request is processed the time of this service request at every turn, and this business process processes the average time of this service request, determines this service request priority of this business process relatively;
According to the priority of every kind of service request, adopt this business process to process each service request that it is assigned with.
7. a business process distributor, is characterized in that, comprising:
Prediction module, for the service request quantity sending according to the user who has received in the time interval of each setting, the service request quantity that in next time interval of prediction current time interval, user sends;
Distribution module, for in the time that the described service request quantity of judgement prediction is greater than the service request quantity that in current time interval, user sends, determine the business of service request quantity and the service request quantity that the interior user of current time interval sends of prediction, according to amassing of this business who determines and current number of processes of having distributed, determine the quantity of the business process increasing, and, according to definite quantity, increase the quantity allotted of business process, in the time that the described service request quantity of judgement prediction is less than the service request quantity that in current time interval, user sends, determine the business of service request quantity and the service request quantity that the interior user of current time interval sends of prediction, and definite predefined parameter and described business's is poor, according to amassing of described difference and current number of processes of having distributed, determine the quantity of the business process reducing, according to definite quantity, reduce the quantity allotted of business process.
8. device as claimed in claim 7, is characterized in that, described prediction module specifically for, adopt difference ARMA model to predict the service request quantity that in next time interval, user sends, comprise and adopt Y
t=α
0+ α
1y
t-1+ ... + α
py
t-p+ β
0μ
t+ β
1μ
t-1+ ... + β
qμ
t-qpredict the service request quantity that in next time interval, user sends, wherein Y
tfor the service request quantity that in next time interval of prediction, user sends, Y
t-1y
t-pfor the service request quantity that user in each this setting-up time interval sends, α
0α
pfor autoregressive coefficient, β
0β
qfor moving average coefficient, μ
tμ
t-qfor rolling average item, the son that p is model returns item number, the rolling average item number that q is model.
9. device as claimed in claim 7, is characterized in that, distribution module also for, the quantity that judges the current business process having distributed is less than the threshold value that business process distributes.
10. device as claimed in claim 7, it is characterized in that, described distribution module specifically for, whether the quantity that judges the business process after increasing is greater than the threshold value that business process distributes, when the quantity of the business process after increasing is greater than the threshold value of business process distribution, the threshold value of distributing according to business process and current number of processes of having distributed poor, redefine the quantity of the business process of increase.
11. devices as described in claim 9 or 10, it is characterized in that, described distribution module also for, add up each business process and process the average time of every kind of service request, for every kind of service request, described service request is assigned to and is processed in the shortest business process described service request average time.
12. devices as claimed in claim 11, it is characterized in that, described distribution module also for, for every kind of service request, the business process being assigned to according to this service request is processed the time of this service request at every turn, and this business process processes the average time of this service request, determines this service request priority of this business process relatively, according to the priority of every kind of service request, adopt this business process to process each service request that it is assigned with.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010595444.6A CN102546203B (en) | 2010-12-20 | 2010-12-20 | Business process allocation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010595444.6A CN102546203B (en) | 2010-12-20 | 2010-12-20 | Business process allocation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102546203A CN102546203A (en) | 2012-07-04 |
| CN102546203B true CN102546203B (en) | 2014-07-02 |
Family
ID=46352215
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010595444.6A Active CN102546203B (en) | 2010-12-20 | 2010-12-20 | Business process allocation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102546203B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103763343A (en) * | 2013-12-27 | 2014-04-30 | 乐视网信息技术（北京）股份有限公司 | Method and device for processing service access |
| CN106161485B (en) * | 2015-03-23 | 2019-03-26 | 腾讯科技（深圳）有限公司 | A kind of resource regulating method of infrastructure service cluster, device and system |
| CN107205248B (en) * | 2016-03-17 | 2021-09-21 | 中国移动通信集团内蒙古有限公司 | Resource allocation method and system |
| CN106375115B (en) * | 2016-08-30 | 2019-12-03 | 东软集团股份有限公司 | Resource allocation methods and device |
| CN109271897B (en) * | 2018-08-31 | 2023-10-24 | 平安科技（深圳）有限公司 | Face action unit detection method and device, computer equipment and storage medium |
| CN110187987B (en) * | 2019-06-05 | 2022-02-25 | 北京百度网讯科技有限公司 | Method and apparatus for processing requests |
| CN110880068A (en) * | 2019-11-08 | 2020-03-13 | 厦门大学 | Method and device for evaluating business handling efficiency, computer equipment and storage medium |
| CN111258678B (en) * | 2020-01-20 | 2022-02-11 | 北京字节跳动网络技术有限公司 | Data allocation method and device, electronic equipment and computer readable medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1404248A (en) * | 2001-09-03 | 2003-03-19 | 华为技术有限公司 | Dynamic channel configuration method based on source rate estimation |
| CN101248619A (en) * | 2005-12-23 | 2008-08-20 | 香港应用科技研究院有限公司 | Distributed wireless network with dynamic bandwidth allocation |
| CN101888342A (en) * | 2010-07-23 | 2010-11-17 | 中兴通讯股份有限公司 | Bandwidth distribution method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070189298A1 (en) * | 2006-02-15 | 2007-08-16 | Hong Kong Applied Science And Technology Research Institute Co., Ltd | Distributed wireless network with dynamic bandwidth allocation |
-
2010
- 2010-12-20 CN CN201010595444.6A patent/CN102546203B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1404248A (en) * | 2001-09-03 | 2003-03-19 | 华为技术有限公司 | Dynamic channel configuration method based on source rate estimation |
| CN101248619A (en) * | 2005-12-23 | 2008-08-20 | 香港应用科技研究院有限公司 | Distributed wireless network with dynamic bandwidth allocation |
| CN101888342A (en) * | 2010-07-23 | 2010-11-17 | 中兴通讯股份有限公司 | Bandwidth distribution method and device |
Non-Patent Citations (2)
| Title |
|---|
| 况忠林.网络流量实时预测与控制.《万方平台学位数据库》.2010,全文. |
| 网络流量实时预测与控制;况忠林;《万方平台学位数据库》;20100729;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102546203A (en) | 2012-07-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102546203B (en) | Business process allocation method and device | |
| US11032212B2 (en) | Systems and methods for provision of a guaranteed batch | |
| EP2193472B1 (en) | Method and apparatus for managing license capacity in a telecommunication network | |
| CN102902587B (en) | Distributed task dispatching mthods, systems and devices | |
| CN111400022A (en) | Resource scheduling method and device and electronic equipment | |
| US9405588B2 (en) | Cloud resource allocation system and method | |
| CN101605362B (en) | Wireless communication system, base station, scheduling method, and program | |
| CN101287299A (en) | Apparatus and method for uplink bandwidth request in broadband wireless access system | |
| CN101248619A (en) | Distributed wireless network with dynamic bandwidth allocation | |
| CN105589750A (en) | CPU (Central Processing Unit) resource scheduling method and server | |
| CN113691596B (en) | Gateway control method and device, electronic equipment and storage medium | |
| CN106161485A (en) | Resource regulating method, device and the system of a kind of infrastructure service cluster | |
| CN106371903A (en) | Task scheduling method for airborne trusted computing platform | |
| CN110099083A (en) | A kind of load equilibration scheduling method and device for server cluster | |
| CN106982245A (en) | Supervise the client and server in Control ＆ data acquisition system | |
| CN114155026A (en) | Resource allocation method, device, server and storage medium | |
| JP2013520896A (en) | Satellite communication resource optimization allocation method and related communication system | |
| US8171484B2 (en) | Resource management apparatus and radio network controller | |
| CN108023936B (en) | Distributed interface access control method and system | |
| CN106411971B (en) | Load adjusting method and device | |
| CN114138453B (en) | Resource optimization allocation method and system suitable for edge computing environment | |
| CN101458631B (en) | Method for scheduling self-adapting virtual machine and computer | |
| Kao et al. | Setting daily production targets with novel approximation of target tracking operations for semiconductor manufacturing | |
| US20140149991A1 (en) | Scheduling system, data processing system, and scheduling method | |
| CN111353712B (en) | Distribution task scheduling method and device, server and storage medium |
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 |