НЕЙРОСЕТЕВАЯ МОДЕЛЬ ПРОГНОЗА ЗАГРЯЗНЕНИЯ АТМОСФЕРНОГО ВОЗДУХА

  • Published on
    05-Apr-2017

  • View
    219

  • Download
    3

Embed Size (px)

Transcript

  • . . 2010. . 4. . 2

    266

    502.7:502.55 .. , . . , ., , (4872) 35-37-60, roeduardo@mail.ru (, , )

    . - DEDUCTOR STUDIO ACADEMIC.

    : , , , , -, , , .

    . - , . . - . , .

    - - , , - , , .

    : ; - ; ; - ; - -.

    , [1]. .1 - (n+1) n - .

    . - x . - . -

  • 267

    . .

    W21

    (1)

    Wn1(1)

    Wn1

    (2)

    . 1. ( )

    -. , , , 5 . - - 5 , .

    x , . . , [0;1]. - - , . , - [0;1]. . .

    5 (X1,X2,X3,X4,X5) , (Y) - - . 1-5 - , Y .

  • . . 2010. . 4. . 2

    268

    - . - - [2], 18 2007 29 2007 . 15832 . Deductor Studio Academic 5.2 BaseGroup Labs, - [3].

    Deductor X1, X2, X3, X4, X5 , Y . 90 10 % .

    0,1 0,9.

    , . - , , .

    0,00508, 0,00381.

    - . : , , .. , - , . , , .

    2 3 2007 , . , . - Deductor . , . .

    . 2 2 3 2007 -, , 18 29 2007 .

  • 269

    .2.

    . 2 1-5

    , Y, Y_OUT Y_ERR. - . , - 0,0064, - 10,7 %.

    .3 176.

    2 3 2007 , -. , 10,7 %, - . - - . -

  • . . 2010. . 4. . 2

    270

    , , , , .

    . 3.

    . - . - , . . - .

    1. ..

    -// . 5. 2010. . 46-52.

    2. . / . .. . .: . 2002. 344 c.

  • 271

    3. BaseGroup. URL.: http://www.basegroup.ru/ ( - 20.01.11).

    E. Roshchupkin Neural-network forecasting model of pollution of atmospheric air The description of result of the forecast of distribution carbon oxigen in atmosphere

    of a city by means of a neural network is resulted. Training of a neural network was made in professional analytical platform deductor studio academic.

    Key words: the forecast, a post of supervision, a time number, monitoring, modeling, a neural network, training, an error.

    02.11.10

    502.7:502.55 .. , . . , ., , (4872) 35-37-60, roeduardo@mail.ru (, , )

    . : , , , ,

    .

    -

    ( -) ( -) [1]. - (). - . - , - , . - . (

Recommended

View more >