Very short-term air pollution forecasting

Patent metadata

Field	Value
Publication number	US-10438125-B2
Application number	US-201514939522-A
Country	US
Kind code	B2
Filing date	Nov 12, 2015
Priority date	Nov 12, 2015
Publication date	Oct 8, 2019
Grant date	Oct 8, 2019

Abstract

A mechanism is provided for forecasting air pollution. One or more air-pollution monitoring stations correlated to a forecasting point from a plurality of air-pollution monitoring stations are identified. For the one or more air-pollution monitoring stations that correlate to the forecasting point, one or more patterns of the forecasting point, historical patterns of the forecasting point relating to the one or more patterns of the forecasting point, and one or more patterns of the air-pollution monitoring stations that relate to the one or more patterns of the forecasting point are identified. Based on the one or more patterns of the forecasting point, the historical patterns of the forecasting point relating to the one or more patterns of the forecasting point, and the one or more patterns of the air-pollution monitoring stations that relate to the one or more patterns of the forecasting point, a pollution forecast is provided.

First claim

What is claimed is:

1. A method, in a data processing system comprising a processor and a memory, the memory comprising instructions that are executed by the processor to cause the processor to be configured to implement an air pollution forecasting mechanism for forecasting air pollution, the method comprising:
identifying, by the air pollution forecasting mechanism, one or more air-pollution monitoring stations correlated to a forecasting point from a plurality of air-pollution monitoring stations, wherein the forecasting point is one of the plurality of air pollution monitoring stations, wherein each of the plurality of air-pollution monitoring stations is located separate from a source of the pollution, and wherein identifying the one or more air-pollution monitoring stations correlated to the forecasting point comprises:
identifying, by the air pollution forecasting mechanism, a set of air-pollution monitoring stations from the plurality of air-pollution monitoring stations that are within a predetermined distance around the forecasting point;
for each of the set of air-pollution monitoring stations, calculating, by the air pollution forecasting mechanism, a diffusion speed s of pollutant from the air-pollution monitoring station M i to the forecasting point M utilizing wind speed w s , and wind direction w d ;
computing, by the air pollution forecasting mechanism, an angle θ between the monitoring station M i and the forecasting point M based on the wind direction w d at site M i ;
computing, by the air pollution forecasting mechanism, a velocity of movement v of the pollutant using the diffusion speed s from the monitoring station M i to the forecasting point M using:
v=s+w s *cos Θ;

identifying, by the air pollution forecasting mechanism, a degree of influence D i of the pollutant from the monitoring station M i to the forecasting point M using:

D
i

=

Q

2
⁢
π
⁢

⁢
v
⁢

⁢

σ
y

⁢

σ
z

⁢

exp
⁡

[

-

1
2

⁢

(

y
2

σ
y
2

⁢
t

+

z
2

σ
z
2

⁢
t

)

-

k
2

⁢
t

]

where Q is the pollution value of M, v is the velocity of movement, ay is the y-axis diffusion parameter (constant value), aσ z is the z-axis diffusion parameter (constant value), y is the y-axis distant between M, and M, z is the z-axis distant between MA and M, t is the lasting hour, and k is the decay factor;
determining, by the air pollution forecasting mechanism, whether the identified degree of influence D i , is greater than a degree of influence threshold DT; and
responsive to the degree of influence D i being greater than a degree of influence threshold DT, adding, by the air pollution forecasting mechanism, the air-pollution monitoring station M i to the one or more air-pollution monitoring stations correlated to the forecasting point M;

for the one or more air-pollution monitoring stations that correlate to the forecasting point, identifying, by the air pollution forecasting mechanism, one or more patterns of the forecasting point, historical patterns of the forecasting point relating to the one or more patterns of the forecasting point, and one or more patterns of the air-pollution monitoring stations that relate to the one or more patterns of the forecasting point; and
providing, by the air pollution forecasting mechanism, a pollution forecast, based on the one or more patterns of the forecasting point, the historical patterns of the forecasting point relating to the one or more patterns of the forecasting point, and the one or more patterns of the air-pollution monitoring stations that relate to the one or more patterns of the forecasting point, to one or more enterprises or individuals in order that action is taken to protect themselves from or to reduce air-pollution,
wherein providing the pollution forecast based on the one or more patterns of the forecasting point, the historical patterns of the forecasting point relating to the one or more patterns of the forecasting point, and the one or more patterns of the air-pollution monitoring stations that relate to the one or more patterns of the forecasting point comprises:
for each of the one or more air-pollution monitoring stations, calculating, by the air pollution forecasting mechanism, a weight w i for a time period Tj according to the similarity Sm at the time period Tj forming Smj and to the similarity Smi at the time period Tj forming Smij using:

w
j

=

Smj
*
Π
⁢

⁢
Smij

∑

j
=
1

n

⁢

Smj

Assignees

• IBM

Inventors

• Bai Xin Xin
• Dong Jin
• Rui Xiao Guang
• Wang hai feng
• Yin Wen Jun

Classifications

• G06N7/00

  Computing arrangements based on specific mathematical models · CPC title

• G06F17/18 Primary

  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 title

• G06N5/047 Primary

  Pattern matching networks; Rete networks · CPC title