Flow rate prediction device, flow rate prediction method, and flow rate prediction program

• US 9,722,892 B2
• Filed: 06/28/2013
• Issued: 08/01/2017
• Est. Priority Date: 07/02/2012
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Yoshida, Hiroshi
• Primary Examiner(s): Acolatse, Kodzovi
• Application Number:
  US14/411,979
• Publication Number:
  US 20150180740A1
• Time in Patent Office: 1,495 Days
• Field of Search: None
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: H04L 41/142 using statistical or mathematical methods
• CPC Class Codes: H04L 43/04 Processing captured monitoring data, e.g. for logfile generation
• CPC Class Codes: H04L 43/0894 Packet rate
• CPC Class Codes: H04L 47/823 Prediction of resource usage
• CPC Class Codes: H04L 47/83 based on usage prediction
• CPC Class Codes: H04W 24/08 Testing, supervising or monitoring using real traffic

Abstract

A flow rate prediction device to predict a flow rate determines, based on time-series data of a measured flow rate, whether the state of the flow rate is stationary or non-stationary. The flow rate prediction device identifies a stochastic process model defining the flow rate as a stochastic variable based on the result of determination of whether the state is stationary or non-stationary and calculates a parameter for use in the identified stochastic process model based on the time-series data. The flow rate prediction device calculates a probability distribution function or a probability density function defining the flow rate to be predicted as the stochastic variable based on the identified stochastic process model and the parameter.

Claims

1. A flow rate prediction device, the device comprising:

a determining unit configured to determine, based on time-series data of a measured flow rate, whether a state of the flow rate is stationary or non-stationary;

an identifying unit configured to

identify a first stochastic process model defining the flow rate as a first stochastic variable if the state of the flow rate is stationary and a second stochastic process model defining the flow rate as a second stochastic variable if the state of the flow rate is non-stationary; and

calculates a first parameter for use in the identified first stochastic process model based on the time-series data and a second parameter for use in the identified second stochastic process model based on the time-series data; and

a calculating unit configured to

calculate a probability distribution function or a probability density function defining a first pedicted flow rate as the first stochastic variable based on the first stochastic process model identified by the identifying unit and the first parameter calculated by the identifying unit, and

calculate a probability distribution function or a probability density function defining a second predicted flow rate as the second stochastic variable based on the second stochastic process model identified by the identifying unit and the second parameter calculated by the identifying unit;

wherein the determining unit, the identifying unit, and the calculating unit are each implemented by at least one hardware processor.

2. The device according to claim 1, further comprising a measuring unit is configured to measure the time-series data of the flow rate.

3. The device according to claim 1, wherein the calculating unit further calculates stochastic diffusion corresponding to stochastic spread of the predicted flow rate based on the probability distribution function or the probability density function.

4. The device according to claim 1, wherein the determining unit performs a unit root test for the determination of whether the state of the measured flow rate is stationary or non-stationary.

5. The device according to claim 1, wherein

the stochastic process model identified by the identifying unit is a non-stationary stochastic process model when the unit determines that the state of the measured flow rate is non-stationary, and

the stochastic process model identified by the identifying unit is a stationary stochastic process model different from the non-stationary stochastic process model when the determining unit determines that the state of the measured flow rate is stationary.

6. The device according to claim 1, wherein the stochastic process model identified by the identifying unit is an autoregression (AR) model.

9. The device according to claim 1,

wherein the probability distribution function or a probability density function is determined by an expected value of the predicted flow rate, and a standard deviation of the flow rate to be predicted.

7. A flow rate prediction method implemented by one or more processors, the method comprising:

determining, based on time-series data of a measured flow rate, whether a state of the flow rate is stationary or non-stationary;

identifying a first stochastic process model defining the flow rate as a first stochastic variable if the state of the flow rate is stationary and a second stochastic process model defining the flow rate as a second stochastic variable if the state of the flow rate is non-stationary;

calculating a first parameter for use in the identified first stochastic process model based on the time-series data and a second parameter for use in the identified second stochastic process model based on the time-series data; and

calculating a probability distribution function or a probability density function defining a first predicted flow rate as the first stochastic variable based on the identified first stochastic process model and the first calculated parameter, and

calculating a probability distribution function or a probability density function defining a second predicted flow rate as the second stochastic variable based on the identified second stochastic process model and the second calculated parameter.

10. The method according to claim 7,

wherein the probability distribution function or a probability density function is determined by an expected value of the predicted flow rate, and a standard deviation of the flow rate to be predicted.

8. A non-transitory computer readable medium embodying instructions for controlling a device to implement a flow rate prediction method, the flow rate prediction method comprising:

determining, based on time-series data of a measured flow rate, whether a state of the flow rate is stationary or non-stationary;

identifying a first stochastic process model defining the flow rate as a first stochastic variable if the state of the flow rate is stationary and a second stochastic process model defining the flow rate as a second stochastic variable if the state of the flow rate is non-stationary;

calculating a first parameter for use in the identified first stochastic process model based on the time-series data and a second parameter for use in the identified second stochastic process model based on the time-series data; and

calculating a probability distribution function or a probability density function defining a first predicted flow rate as the first stochastic variable based on the identified first stochastic process model and the first calculated parameter, and

calculating a probability distribution function or a probability density function defining a second predicted flow rate as the second stochastic variable based on the identified second stochastic process model and the second calculated parameter.

11. The non-transitory computer readable medium according to claim 8,

wherein the probability distribution function or a probability density function is determined by an expected value of the predicted flow rate, and a standard deviation of the flow rate to be predicted.