close

Вход

Забыли?

вход по аккаунту

...весна — 2015»: итоги конкурса инструментальной музыки;pdf

код для вставкиСкачать
ɍȾɄ 004.9
ȼ. Ⱥ. ȿɦɟɥɶɹɧɨɜ, ɝ.ɋɟɜɚɫɬɨɩɨɥɶ
ɆɈȾȿɅɂɊɈȼȺɇɂȿ ɆɇɈȽɈɋɅɈɃɇɕɏ ɇȿɃɊɈɇɇɕɏ ɋȿɌȿɃ ȼ
MATLAB ȾɅə ɊȺɋɉɈɁɇȺȼȺɇɂə ɂɁɈȻɊȺɀȿɇɂɃ
ɆɂɄɊɈɋɌɊɍɄɌɍɊ ɆȿɌȺɅɅɈȼ
Abstract. The urgency to develop new methods and means of automated
metallographic analysis with the possibility to diagnose of new metals has been
substantiated. The usage of the neural networks to diagnose the metals state has been
proposed. The multilayer neural networks in Matlab have been modeling. The lowest
mean square error of the proposed multilayer neural networks for image recognition of
the metals microstructures has been obtained.
Ⱥɤɬɭɚɥɶɧɨɫɬɶ
Ƚɥɨɛɚɥɢɡɚɰɢɹ ɬɟɯɧɢɱɟɫɤɨɣ ɞɢɚɝɧɨɫɬɢɤɢ ɩɪɟɞɭɫɦɚɬɪɢɜɚɟɬ ɭɜɟɥɢɱɟɧɢɟ
ɤɨɥɢɱɟɫɬɜɚ ɞɢɚɝɧɨɫɬɢɱɟɫɤɢɯ ɨɩɟɪɚɰɢɣ ɢ ɬɟɯɧɨɥɨɝɢɣ ɞɥɹ ɤɨɧɬɪɨɥɹ ɤɚɱɟɫɬɜɚ ɢ
ɬɟɯɧɢɱɟɫɤɨɝɨ ɫɨɫɬɨɹɧɢɹ ɦɟɬɚɥɥɨɜ ɢ ɢɡɞɟɥɢɣ ɞɥɹ ɞɨɫɬɢɠɟɧɢɹ ɬɟɯɧɨɝɟɧɧɨɣ
ɛɟɡɨɩɚɫɧɨɫɬɢ, ɤɨɬɨɪɚɹ ɜ ɫɜɨɸ ɨɱɟɪɟɞɶ ɬɪɟɛɭɟɬ ɪɚɡɪɚɛɨɬɤɢ ɬɟɯɧɢɱɟɫɤɢɯ
ɫɪɟɞɫɬɜ ɢ ɢɧɮɨɪɦɚɰɢɨɧɧɵɯ ɬɟɯɧɨɥɨɝɢɣ. Ⱦɥɹ ɪɟɲɟɧɢɹ ɩɪɨɛɥɟɦɵ ɬɟɯɧɢɱɟɫɤɨɣ
ɞɢɚɝɧɨɫɬɢɤɢ ɦɟɬɚɥɥɨɜ, ɢɫɩɨɥɶɡɭɸɬɫɹ ɨɛɴɟɤɬɢɜɧɵɟ ɮɢɡɢɱɟɫɤɢɟ ɦɟɬɨɞɵ
ɤɨɧɬɪɨɥɹ ɫɨɫɬɨɹɧɢɹ ɢɡɞɟɥɢɹ, ɬɚɤɢɟ ɤɚɤ ɦɟɬɨɞɵ ɧɟɪɚɡɪɭɲɚɸɳɟɝɨ ɤɨɧɬɪɨɥɹ ɢ
ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɢɣ ɚɧɚɥɢɡ [1]. ɋɭɬɶ ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɨɝɨ ɚɧɚɥɢɡɚ ɫɜɨɞɢɬɫɹ
ɤ ɞɢɚɝɧɨɫɬɢɤɟ ɫɨɫɬɨɹɧɢɹ ɦɟɬɚɥɥɨɜ ɧɚ ɨɫɧɨɜɟ ɪɚɫɩɨɡɧɚɜɚɧɢɹ ɢɡɨɛɪɚɠɟɧɢɣ ɢɯ
ɦɢɤɪɨɫɬɪɭɤɬɭɪ.
ɉɨɫɬɚɧɨɜɤɚ ɡɚɞɚɱɢ
ɋɪɟɞɢ ɧɚɢɛɨɥɟɟ ɡɧɚɱɢɬɟɥɶɧɵɯ ɪɚɛɨɬ ɜ ɨɛɥɚɫɬɢ ɚɜɬɨɦɚɬɢɡɚɰɢɢ
ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɨɝɨ ɚɧɚɥɢɡɚ ɦɨɠɧɨ ɜɵɞɟɥɢɬɶ ɪɚɛɨɬɵ ɨɬɟɱɟɫɬɜɟɧɧɵɯ
ɭɱɟɧɵɯ Ⱥ.ɋ. əɤɨɜɥɟɜɚ, Ɇ.ɋ. Ɏɢɥɢɧɨɜɚ, Ⱥ.Ⱥ. ɑɭɛɨɜɚ, ɚ ɬɚɤɠɟ ɪɚɛɨɬɵ
ɡɚɪɭɛɟɠɧɵɯ ɭɱɟɧɵɯ, ɪɟɡɭɥɶɬɚɬɵ, ɪɚɛɨɬɵ ɤɨɬɨɪɵɯ ɢɡɥɨɠɟɧɵ ɜ ɬɪɭɞɚɯ [2,3].
Ʉɪɨɦɟ ɬɨɝɨ, ɚɜɬɨɦɚɬɢɡɚɰɢɟɣ ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɨɝɨ ɚɧɚɥɢɡɚ ɜ ɧɚɫɬɨɹɳɟɟ ɜɪɟɦɹ
ɡɚɧɢɦɚɟɬɫɹ ɪɹɞ ɡɚɪɭɛɟɠɧɵɯ ɢ ɨɬɟɱɟɫɬɜɟɧɧɵɯ ɮɢɪɦ: "SIAMS", "VideoTest",
“VisionPE” [4-6].
Ɉɞɧɚɤɨ ɛɨɥɶɲɢɧɫɬɜɨ ɫɭɳɟɫɬɜɭɸɳɢɯ ɪɟɲɟɧɢɣ ɜ ɞɚɧɧɨɣ ɨɛɥɚɫɬɢ
ɨɪɢɟɧɬɢɪɨɜɚɧɵ ɧɚ ɨɩɪɟɞɟɥɟɧɧɵɟ ɜɢɞɵ ɦɟɬɚɥɥɨɜ ɢ ɜ ɷɬɢɯ ɫɪɟɞɫɬɜɚɯ
ɨɬɫɭɬɫɬɜɭɟɬ ɜɨɡɦɨɠɧɨɫɬɶ ɨɛɭɱɟɧɢɹ ɫɢɫɬɟɦɵ ɞɢɚɝɧɨɫɬɢɤɟ ɧɨɜɵɯ ɦɚɪɨɤ
ɦɟɬɚɥɥɨɜ. Ɍɚɤɢɦ ɨɛɪɚɡɨɦ, ɚɤɬɭɚɥɶɧɨɣ ɹɜɥɹɟɬɫɹ ɡɚɞɚɱɚ ɪɚɡɪɚɛɨɬɤɢ ɧɨɜɵɯ
ɚɥɝɨɪɢɬɦɢɱɟɫɤɢɯ ɢ ɢɧɫɬɪɭɦɟɧɬɚɥɶɧɵɯ ɫɪɟɞɫɬɜ ɚɜɬɨɦɚɬɢɡɢɪɨɜɚɧɧɨɝɨ
ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɨɝɨ ɚɧɚɥɢɡɚ ɫ ɜɨɡɦɨɠɧɨɫɬɶɸ ɢɯ ɨɛɭɱɟɧɢɹ ɞɢɚɝɧɨɫɬɢɤɟ
ɧɨɜɵɯ ɦɚɪɨɤ ɦɟɬɚɥɥɨɜ
Ɋɟɲɟɧɢɟ ɡɚɞɚɱɢ
ɂɡɨɛɪɚɠɟɧɢɟ ɦɢɤɪɨɫɬɪɭɤɬɭɪɵ ɦɟɬɚɥɥɨɜ ɮɨɪɦɢɪɭɟɬɫɹ ɩɨɫɪɟɞɫɬɜɨɦ
ɫɩɟɰɢɚɥɶɧɨɝɨ
ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɨɝɨ ɦɢɤɪɨɫɤɨɩɚ, ɩɨɞɤɥɸɱɟɧɧɨɝɨ
ɤ
© ȼ. Ⱥ. ȿɦɟɥɶɹɧɨɜ
37
ɤɨɦɩɶɸɬɟɪɭ. ɉɨɥɭɱɟɧɧɨɟ ɩɨɥɭɬɨɧɨɜɨɟ ɢɡɨɛɪɚɠɟɧɢɟ ɦɢɤɪɨɫɬɪɭɤɬɭɪɵ ɦɟɬɚɥɥɚ
f(x,y) ɩɨɞɜɟɪɝɚɟɬɫɹ ɛɢɧɚɪɢɡɚɰɢɢ.
Ⱦɥɹ ɛɢɧɚɪɢɡɚɰɢɢ ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɢɯ ɢɡɨɛɪɚɠɟɧɢɣ ɩɪɟɞɥɚɝɚɟɬɫɹ
ɢɫɩɨɥɶɡɨɜɚɧɢɟ ɦɟɬɨɞɚ ɫ ɢɡɦɟɧɹɸɳɢɦɫɹ ɩɨɪɨɝɨɦ ɩɪɟɨɛɪɚɡɨɜɚɧɢɹ [7], ɷɬɨ
ɨɛɭɫɥɨɜɥɟɧɨ ɬɟɦ, ɱɬɨ ɭɱɢɬɵɜɚɬɶ ɧɚɣɞɟɧɧɵɟ ɩɨɪɨɝɢ ɞɥɹ ɛɥɢɠɚɣɲɢɯ ɭɱɚɫɬɤɨɜ
ɧɟɨɛɯɨɞɢɦɨ ɞɥɹ ɦɢɧɢɦɢɡɚɰɢɢ ɜɨɡɦɨɠɧɵɯ ɨɲɢɛɨɤ ɜ ɦɟɫɬɚɯ ɫ ɫɢɥɶɧɵɦɢ
ɞɟɮɟɤɬɚɦɢ ɢɡɨɛɪɚɠɟɧɢɹ ɦɢɤɪɨɫɬɪɭɤɬɭɪ.
ɉɨɫɥɟ ɛɢɧɚɪɢɡɚɰɢɢ ɧɟɨɛɯɨɞɢɦɨ ɨɩɪɟɞɟɥɢɬɶ ɢɧɮɨɪɦɚɬɢɜɧɵɟ ɩɪɢɡɧɚɤɢ
ɢɡɨɛɪɚɠɟɧɢɹ. Ʉɚɤ ɢɡɜɟɫɬɧɨ [7,8], c ɬɨɱɤɢ ɡɪɟɧɢɹ ɪɚɫɩɨɡɧɚɜɚɧɢɹ ɢ ɚɧɚɥɢɡɚ
ɨɛɴɟɤɬɨɜ ɧɚ ɢɡɨɛɪɚɠɟɧɢɢ ɧɚɢɛɨɥɟɟ ɢɧɮɨɪɦɚɬɢɜɧɵɦɢ ɹɜɥɹɸɬɫɹ ɧɟ ɡɧɚɱɟɧɢɹ
ɹɪɤɨɫɬɟɣ ɨɛɴɟɤɬɨɜ, ɚ ɯɚɪɚɤɬɟɪɢɫɬɢɤɢ ɢɯ ɝɪɚɧɢɰ – ɤɨɧɬɭɪɨɜ. Ⱦɪɭɝɢɦɢ ɫɥɨɜɚɦɢ,
ɨɫɧɨɜɧɚɹ ɢɧɮɨɪɦɚɰɢɹ ɡɚɤɥɸɱɟɧɚ ɧɟ ɜ ɹɪɤɨɫɬɢ ɨɬɞɟɥɶɧɵɯ ɨɛɥɚɫɬɟɣ, ɚ ɜ ɢɯ
ɨɱɟɪɬɚɧɢɹɯ. Ɍɚɤɢɦ ɨɛɪɚɡɨɦ, ɡɚɞɚɱɚ ɜɵɞɟɥɟɧɢɹ ɤɨɧɬɭɪɨɜ ɫɨɫɬɨɢɬ ɜ ɩɨɫɬɪɨɟɧɢɢ
ɢɡɨɛɪɚɠɟɧɢɹ ɢɦɟɧɧɨ ɝɪɚɧɢɰ ɨɛɴɟɤɬɨɜ ɢ ɨɱɟɪɬɚɧɢɣ ɨɞɧɨɪɨɞɧɵɯ ɨɛɥɚɫɬɟɣ.
ɋɥɟɞɨɜɚɬɟɥɶɧɨ, ɞɚɥɟɟ ɩɪɨɢɡɜɨɞɢɬɫɹ ɜɵɞɟɥɟɧɢɟ ɝɪɚɧɢɰ ɫɟɝɦɟɧɬɨɜ ɢ
ɨɩɪɟɞɟɥɟɧɢɟ ɛɚɡɨɜɵɯ ɬɨɱɟɤ ɢɡɨɛɪɚɠɟɧɢɹ ɮɢɥɶɬɪɨɦ ɉɪɟɜɢɬɬɚ (ɤɚɤ ɧɚɢɛɨɥɟɟ
ɩɨɦɟɯɨɭɫɬɨɣɱɢɜɵɦ ɮɢɥɶɬɪɨɦ ɫɨɝɥɚɫɧɨ [7,8]).
ɗɥɟɦɟɧɬɵ ɫɟɝɦɟɧɬɚ ɮɨɪɦɢɪɭɸɬɫɹ ɩɪɢ ɩɨɦɨɳɢ ɝɢɩɨɬɟɧɭɡ ɬɪɟɭɝɨɥɶɧɢɤɨɜ,
ɤɨɬɨɪɵɟ ɮɨɪɦɢɪɭɸɬɫɹ ɩɨɫɪɟɞɫɬɜɨɦ ɩɟɪɩɟɧɞɢɤɭɥɹɪɨɜ, ɨɩɭɳɟɧɧɵɯ ɢɡ ɞɜɭɯ
ɫɨɫɟɞɧɢɯ ɛɚɡɨɜɵɯ ɬɨɱɟɤ, ɤɚɤ ɩɪɨɢɥɥɸɫɬɪɢɪɨɜɚɧɨ ɧɚ ɪɢɫɭɧɤɟ 1.
Ɋɢɫ. 1. ɋɟɝɦɟɧɬ ɚɧɚɥɢɡɢɪɭɟɦɨɝɨ ɢɡɨɛɪɚɠɟɧɢɹ ɦɢɤɪɨɫɬɪɭɤɬɭɪɵ ɦɟɬɚɥɥɚ
Ɉɩɢɫɵɜɚɟɬɫɹ, ɩɪɨɢɥɥɸɫɬɪɢɪɨɜɚɧɧɵɣ ɫɟɝɦɟɧɬ ɡɧɚɱɟɧɢɹɦɢ ɮɭɧɤɰɢɣ sin ɢ
cos ɭɝɥɨɜ Ⱥ ɢ ȼ. Ɍɚɤɢɦ ɨɛɪɚɡɨɦ, ɢɧɮɨɪɦɚɰɢɨɧɧɵɦɢ ɯɚɪɚɤɬɟɪɢɫɬɢɤɚɦɢ
ɢɡɨɛɪɚɠɟɧɢɹɦɢ ɦɢɤɪɨɫɬɪɭɤɬɭɪɵ ɦɟɬɚɥɥɚ ɹɜɥɹɟɬɫɹ ɦɧɨɠɟɫɬɜɨ:
ɉi
^sin Ai , cos Ai , Gpi ` .
(1)
Ⱦɚɥɟɟ ɪɚɫɩɨɡɧɚɜɚɧɢɟ ɩɪɟɞɥɚɝɚɟɬɫɹ ɩɪɨɜɨɞɢɬɶ ɫ ɩɨɦɨɳɶɸ ɧɟɣɪɨɧɧɵɯ
ɫɟɬɟɣ, ɱɬɨ ɩɨɡɜɨɥɢɬ ɪɚɡɪɚɛɚɬɵɜɚɟɦɵɦ ɫɪɟɞɫɬɜɚɦ ɩɪɢɞɚɬɶ ɫɜɨɣɫɬɜɨ
ɨɛɭɱɚɟɦɨɫɬɢ ɞɥɹ ɩɪɟɞɨɫɬɚɜɥɟɧɢɹ ɜɨɡɦɨɠɧɨɫɬɢ ɞɢɚɝɧɨɫɬɢɤɢ ɫɨɫɬɨɹɧɢɹ
ɪɚɡɧɵɯ ɦɚɪɨɤ ɦɟɬɚɥɥɨɜ.
ɇɚɛɨɪ ɜɫɟɯ ɡɧɚɱɟɧɢɣ ɮɭɧɤɰɢɣ sin(A), cos(A) ɢ ɝɪɚɞɢɟɧɬɚ ɉɪɟɜɢɬɬɚ Gp,
ɤɨɬɨɪɵɟ ɯɚɪɚɤɬɟɪɢɡɭɸɬ ɛɚɡɨɜɵɟ ɬɨɱɤɢ ɢɡɨɛɪɚɠɟɧɢɹ, ɢ ɟɝɨ ɫɟɝɦɟɧɬɵ ɹɜɥɹɟɬɫɹ
ɜɯɨɞɧɵɦ ɫɢɝɧɚɥɨɦ ɞɥɹ ɦɨɞɟɥɢɪɭɟɦɵɯ ɧɟɣɪɨɧɧɵɯ ɫɟɬɟɣ.
38
ɋɬɪɭɤɬɭɪɚ ɦɧɨɝɨɫɥɨɣɧɨɣ ɧɟɣɪɨɧɧɨɣ ɫɟɬɢ ɨɩɪɟɞɟɥɟɧɢɹ ɤɨɥɢɱɟɫɬɜɟɧɧɵɯ
ɯɚɪɚɤɬɟɪɢɫɬɢɤ ɫɬɚɥɟɣ, ɬɚɤɢɯ ɤɚɤ ɛɚɥɥ ɡɟɪɧɚ, ɩɪɢɜɟɞɟɧɚ ɜ ɪɚɛɨɬɟ [9].
ɉɪɟɞɥɚɝɚɟɬɫɹ ɤɨɥɢɱɟɫɬɜɨ ɧɟɣɪɨɧɨɜ ɜɨ ɜɯɨɞɧɨɦ ɫɥɨɟ ɩɪɢɧɢɦɚɬɶ ɪɚɜɧɵɦ
ɫɪɟɞɧɟɦɭ ɤɨɥɢɱɟɫɬɜɭ ɛɚɡɨɜɵɯ ɬɨɱɟɤ ɧɚ ɢɡɨɛɪɚɠɟɧɢɹɯ ɦɢɤɪɨɫɬɪɭɤɬɭɪ
ɦɟɬɚɥɥɨɜ ɞɥɹ ɡɚɞɚɧɧɨɝɨ ɬɢɩɚ ɦɟɬɚɥɥɨɝɪɚɮɢɱɟɫɤɨɝɨ ɚɧɚɥɢɡɚ.
ȼɟɥɢɱɢɧɚ ɫɤɪɵɬɨɝɨ ɫɥɨɹ ɜɵɱɢɫɥɹɟɬɫɹ ɞɟɥɟɧɢɟɦ ɧɚ 3 ɤɨɥɢɱɟɫɬɜɚ
ɧɟɣɪɨɧɨɜ ɜɯɨɞɧɨɝɨ ɫɥɨɹ, ɩɨɫɤɨɥɶɤɭ ɜ ɫɤɪɵɬɨɦ ɫɥɨɟ ɩɪɨɢɡɜɨɞɢɬɫɹ
ɫɟɝɦɟɧɬɚɰɢɹ ɢɡɨɛɪɚɠɟɧɢɹ ɧɚ ɨɫɧɨɜɟ ɛɚɡɨɜɵɯ ɬɨɱɟɤ ɫɟɝɦɟɧɬɨɜ, ɬ.ɟ. ɞɥɹ ɤɚɠɞɨɣ
ɬɨɱɤɢ ɫɟɝɦɟɧɬɚ ɜɵɱɢɫɥɹɟɬɫɹ ɯɚɪɚɤɬɟɪɢɡɭɸɳɟɟ ɟɺ ɚɩɩɪɨɤɫɢɦɢɪɭɟɦɨɟ
ɡɧɚɱɟɧɢɟ ɧɚ ɨɫɧɨɜɟ ɩɚɪɚɦɟɬɪɨɜ ɛɚɡɨɜɵɯ ɬɨɱɟɤ.
Ɋɚɡɦɟɪ ɜɵɯɨɞɧɨɝɨ ɫɥɨɹ ɨɩɪɟɞɟɥɹɟɬɫɹ ɤɨɥɢɱɟɫɬɜɨɦ ɦɚɪɨɤ ɦɟɬɚɥɥɨɜ ɞɥɹ
ɪɚɫɩɨɡɧɚɜɚɧɢɹ.
ɉɨɫɥɟ ɜɵɛɨɪɚ ɫɬɪɭɤɬɭɪɵ ɧɟɣɪɨɧɧɨɣ ɫɟɬɢ ɧɟɨɛɯɨɞɢɦɨ ɟɟ ɨɛɭɱɢɬɶ. Ⱦɥɹ
ɪɟɲɟɧɢɹ ɛɵɥ ɜɵɛɪɚɧ ɚɥɝɨɪɢɬɦ ɨɛɪɚɬɧɨɝɨ ɪɚɫɩɪɨɫɬɪɚɧɟɧɢɹ ɨɲɢɛɤɢ
(backpropagation) [10]. Ⱦɥɹ ɚɥɝɨɪɢɬɦɚ ɨɛɪɚɬɧɨɝɨ ɪɚɫɩɪɨɫɬɪɚɧɟɧɢɹ
ɧɟɨɛɯɨɞɢɦɨ, ɱɬɨɛɵ ɚɤɬɢɜɚɰɢɨɧɧɚɹ ɮɭɧɤɰɢɹ ɛɵɥɚ ɞɢɮɮɟɪɟɧɰɢɪɭɟɦɚ ɧɚ ɜɫɟɦ
ɭɱɚɫɬɤɟ.
ɉɪɟɞɥɚɝɚɟɬɫɹ ɢɫɩɨɥɶɡɨɜɚɬɶ ɫɢɝɦɨɢɞɚɥɶɧɭɸ ɚɤɬɢɜɚɰɢɨɧɧɭɸ ɮɭɧɤɰɢɸ,
ɩɨɫɤɨɥɶɤɭ ɨɧɚ ɭɞɨɜɥɟɬɜɨɪɹɟɬ ɞɚɧɧɨɦɭ ɭɫɥɨɜɢɸ [10]:
yi
1
1 e xi
.
(2)
Ɉɛɭɱɟɧɢɟ ɧɟɣɪɨɧɧɨɣ ɫɟɬɢ ɩɪɨɢɡɜɨɞɢɥɨɫɶ ɧɚ ɨɫɧɨɜɟ ɷɬɚɥɨɧɧɵɯ
ɢɡɨɛɪɚɠɟɧɢɣ ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɫɩɥɚɜɨɜ ɨɩɢɫɚɧɧɵɯ ɜ ȽɈɋɌ 5639-82, ɜ ȽɈɋɌ
1778-70, ȽɈɋɌ 8233-56 ɢ ɞɪɭɝɢɯ ɫɬɚɧɞɚɪɬɚɯ:
Ɋɢɫ. 2. ɉɪɢɦɟɪ ɨɛɭɱɚɸɳɟɣ ɜɵɛɨɪɤɢ ɢɡɨɛɪɚɠɟɧɢɣ ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɫɨɝɥɚɫɧɨ ȽɈɋɌ
8233-56
Ɉɛɭɱɚɸɳɚɹ ɜɵɛɨɪɤɚ ɫɨɫɬɚɜɢɥɚ 280 ɢɡɨɛɪɚɠɟɧɢɣ ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɫɬɚɥɟɣ
ɦɚɪɨɤ 10ɏɋɇȾ, 20ɏȽɋȺ, 30ɏȽɌ, ɩɪɢ ɷɬɨɦ, ɢɡ ɧɢɯ 140 «ɯɨɪɨɲɢɯ» ɢ 140
«ɩɥɨɯɢɯ». ɉɨɞ «ɯɨɪɨɲɢɦɢ» ɩɨɧɢɦɚɸɬɫɹ ɢɡɨɛɪɚɠɟɧɢɹ ɷɬɚɥɨɧɧɵɯ
ɦɢɤɪɨɫɬɪɭɤɬɭɪ, ɚ ɩɨɞ «ɩɥɨɯɢɦɢ» ɩɪɢɦɟɪɵ ɢɫɤɚɠɟɧɧɵɯ ɲɭɦɚɦɢ ɢɡɨɛɪɚɠɟɧɢɣ
ɷɬɚɥɨɧɨɜ, ɱɬɨ ɤɚɤ ɫɥɟɞɫɬɜɢɟ ɜɟɞɟɬ ɤ ɧɟɩɪɚɜɢɥɶɧɨɦɭ ɪɚɫɩɨɡɧɚɜɚɧɢɸ
39
(ɤɥɚɫɫɢɮɢɤɚɰɢɢ) ɢɡɨɛɪɚɠɟɧɢɹ ɧɟɣɪɨɧɧɨɣ ɫɟɬɶɸ. Ɍɚɤɢɦ ɨɛɪɚɡɨɦ, ɧɟɣɪɨɧɧɚɹ
ɫɟɬɶ ɬɚɤɠɟ ɨɛɭɱɚɥɚɫɶ ɧɟɜɟɪɧɨɦɭ ɪɚɫɩɨɡɧɚɜɚɧɢɸ, ɬ.ɟ. ɪɟɚɝɢɪɨɜɚɧɢɸ ɧɚ
ɧɟɤɨɪɪɟɤɬɧɵɟ ɢɡɨɛɪɚɠɟɧɢɹ.
ȼ ɤɚɱɟɫɬɜɟ ɤɨɧɬɪɨɥɶɧɨɣ ɜɵɛɨɪɤɢ ɢɫɩɨɥɶɡɨɜɚɥɨɫɶ 180 ɢɡɨɛɪɚɠɟɧɢɣ
ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɫɩɥɚɜɨɜ:
Ɋɢɫ. 3. ɉɪɢɦɟɪ ɤɨɧɬɪɨɥɶɧɨɣ ɜɵɛɨɪɤɢ ɢɡɨɛɪɚɠɟɧɢɣ ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɢɡ ɛɚɡɵ
ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɦɟɬɚɥɥɨɜ ɢ ɫɩɥɚɜɨɜ [11]
ɉɪɢ ɦɨɞɟɥɢɪɨɜɚɧɢɢ ɩɪɟɞɥɨɠɟɧɧɨɣ ɫɬɪɭɤɬɭɪɵ ɧɟɣɪɨɧɧɨɣ ɫɟɬɢ ɜ Matlab
Ɇ-ɮɭɧɤɰɢɹ ɮɨɪɦɢɪɭɟɬ ɜɵɯɨɞɧɵɟ ɩɟɪɟɦɟɧɧɵɟ input ɢ targets, ɤɨɬɨɪɵɟ
ɨɩɪɟɞɟɥɹɸɬ ɦɚɫɫɢɜɵ ɜɯɨɞɧɵɯ ɡɧɚɱɟɧɢɣ ɯɚɪɚɤɬɟɪɢɫɬɢɤ ɢɡɨɛɪɚɠɟɧɢɣ
ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɦɟɬɚɥɥɨɜ ɉi ^sin Ai , cos Ai , Gpi ` ɢ ɰɟɥɟɜɵɯ ɜɟɤɬɨɪɨɜ.
ɋɨɡɞɚɧɢɟ ɧɟɣɪɨɧɧɵɯ ɫɟɬɟɣ ɩɨ ɫɩɪɨɟɤɬɢɪɨɜɚɧɧɨɣ ɫɬɪɭɤɬɭɪɟ [9]
ɨɫɭɳɟɫɬɜɥɹɥɨɫɶ ɩɪɢ ɩɨɦɨɳɢ ɤɨɦɚɧɞɵ MATLAB – newff.
net = newff([0 1],[150 1],{'logsig','logsig'},'traingd');
gensim(net)
Ɉɛɭɱɟɧɢɟ ɫɟɬɟɣ ɩɪɨɜɨɞɢɥɨɫɶ, ɫ ɡɚɞɚɧɢɟɦ ɤɨɥɢɱɟɫɬɜɚ ɷɩɨɯ ɨɛɭɱɟɧɢɹ:
net.trainParam.epochs = 1000;
net = train(net,input,Targets);
Ɇɨɞɟɥɢɪɨɜɚɧɢɟ ɧɟɣɪɨɧɧɵɯ ɫɟɬɟɣ ɢ ɩɨɫɬɪɨɟɧɢɟ ɝɪɚɮɢɤɨɜ ɫɢɝɧɚɥɚ ɜɵɯɨɞɚ
ɢ ɰɟɥɢ ɩɪɨɜɨɞɢɥɨɫɶ ɤɨɦɚɧɞɚɦɢ:
Y = sim(net,input); figure(1), clf
plot(input, Targets, input, Y, 'o')
ɉɪɢ ɷɬɨɦ, ɩɨɫɤɨɥɶɤɭ ɧɟɨɛɯɨɞɢɦɨ, ɱɬɨɛɵ ɧɟɣɪɨɧɧɚɹ ɫɟɬɶ ɨɛɭɱɚɥɚɫɶ ɧɟ
ɡɚɩɨɦɢɧɚɧɢɸ ɨɞɧɨɝɨ ɨɬɞɟɥɶɧɨɝɨ ɨɛɪɚɡɚ ɦɟɬɚɥɥɚ, ɚ ɨɛɭɱɚɥɚɫɶ ɨɛɨɛɳɟɧɢɸ, ɬɨ
ɛɵɥɢ ɩɪɨɜɟɞɟɧɵ ɷɤɫɩɟɪɢɦɟɧɬɵ ɫ ɪɚɡɧɵɦɢ ɤɨɥɢɱɟɫɬɜɚɦɢ ɜɯɨɞɧɵɯ ɢ
ɫɨɨɬɜɟɬɫɬɜɟɧɧɨ ɫɤɪɵɬɵɯ ɧɟɣɪɨɧɨɜ. Ʉɪɨɦɟ ɬɨɝɨ, ɜ ɢɫɫɥɟɞɨɜɚɧɢɢ ɩɪɨɜɟɞɟɧɵ
ɷɤɫɩɟɪɢɦɟɧɬɵ ɫɨ ɫɥɟɞɭɸɳɢɦɢ Ɇ-ɮɭɧɤɰɢɹɦɢ, ɤɨɬɨɪɵɟ ɩɨɡɜɨɥɹɸɬ ɨɛɭɱɚɬɶ
ɧɟɣɪɨɧɧɭɸ ɫɟɬɶ ɩɨ ɪɚɡɧɵɦ ɚɥɝɨɪɢɬɦɚɦ:
40
- m-ɮɭɧɤɰɢɹ cgb – ɦɟɬɨɞ ɫɜɹɡɚɧɧɵɯ ɝɪɚɞɢɟɧɬɨɜ ɉɚɭɷɥɥɚ-Ȼɢɥɚ;
- m-ɮɭɧɤɰɢɹ cgf – ɦɟɬɨɞ ɫɜɹɡɚɧɧɵɯ ɝɪɚɞɢɟɧɬɨɜ Ɏɥɟɬɱɟɪɚ-ɉɚɭɷɥɥɚ;
- m-ɮɭɧɤɰɢɹ cgp – ɦɟɬɨɞ ɫɜɹɡɚɧɧɵɯ ɝɪɚɞɢɟɧɬɨɜ ɉɨɥɚɤɚ-Ɋɢɛɢɪɚ;
- m-ɮɭɧɤɰɢɹ gd – ɦɟɬɨɞ ɝɪɚɞɢɟɧɬɧɨɝɨ ɫɩɭɫɤɚ;
- m-ɮɭɧɤɰɢɹ gda – ɦɟɬɨɞ ɝɪɚɞɢɟɧɬɧɨɝɨ ɫɩɭɫɤɚ ɫ ɚɞɚɩɬɢɜɧɵɦ ɨɛɭɱɟɧɢɟɦ
(ɫ ɤɨɪɪɟɤɰɢɟɣ ɲɚɝɚ ɨɛɭɱɟɧɢɹ);
Ɏɭɧɤɰɢɹ MATLAB logsig ɦɨɞɟɥɢɪɭɟɬ ɫɢɝɦɨɢɞɚɥɶɧɭɸ ɮɭɧɤɰɢɸ
ɚɤɬɢɜɚɰɢɢ (ɮɨɪɦɭɥɚ 2).
Ⱦɥɹ ɢɡɦɟɪɟɧɢɹ ɤɚɱɟɫɬɜɚ ɪɚɫɩɨɡɧɚɜɚɧɢɹ ɩɪɨɢɡɜɨɞɢɥɨɫɶ ɜɵɱɢɫɥɟɧɢɟ
ɫɪɟɞɧɟɤɜɚɞɪɚɬɢɱɟɫɤɨɣ ɨɲɢɛɤɢ ɩɨ ɮɨɪɦɭɥɟ:
E
1 n
( yi y (ki )) 2 ,
ni 1
¦
(3)
ɝɞɟ E – ɨɲɢɛɤɚ ɪɚɫɩɨɡɧɚɜɚɧɢɹ;
yi – ɡɧɚɱɟɧɢɟ i-ɝɨ ɜɵɯɨɞɚ ɫɟɬɢ ɩɪɢ ɪɚɫɩɨɡɧɚɜɚɧɢɢ ɢɡɨɛɪɚɠɟɧɢɹ;
y (ki ) – ɡɧɚɱɟɧɢɟ i-ɝɨ ɷɬɚɥɨɧɧɨɝɨ ɜɵɯɨɞɚ ɫɟɬɢ, ɤɨɬɨɪɨɟ ɫɨɨɬɜɟɬɫɬɜɭɟɬ ɤ
ɤɥɚɫɫɭ ɢɡɨɛɪɚɠɟɧɢɹ.
Ɋɟɡɭɥɶɬɚɬɵ ɷɤɫɩɟɪɢɦɟɧɬɨɜ ɩɪɨɜɟɞɟɧɧɵɯ ɜ ɫɪɟɞɟ MATLAB, ɫɜɟɞɟɧɵ ɜ
ɬɚɛɥɢɰɭ 1.
Ⱥɥɝɨɪ
ɢɬɦ
gd
gd
gd
gd
cgf
cgf
cgf
cgf
cgp
cgp
cgp
cgp
cgb
cgb
cgb
cgb
gda
gda
gda
gda
Ɍɚɛɥɢɰɚ 1
Ɋɟɡɭɥɶɬɚɬɵ ɷɤɫɩɟɪɢɦɟɧɬɨɜ ɩɪɨɜɟɞɟɧɧɵɯ ɜ ɫɪɟɞɟ MATLAB
ɋɬɪɭɤɬɭɪɚ
ɗɩɨɯɢ
Ɋɚɫɩɨɡɧɚɜɚɧɢɟ
Ɋɚɫɩɨɡɧɚɜɚɧɢɟ
ɧɟɣɪɨɧɧɨɣ
ɨɛɭɱɟɧ
(ɨɛɭɱɚɸɳɚɹ
(ɤɨɧɬɪɨɥɶɧɚɹ
ɫɟɬɢ
ɢɹ
ɜɵɛɨɪɤɚ)
ɜɵɛɨɪɤɚ)
Ok
Error
Ok
Error
810-270-10
1000
93.9
0.0
93.5
1.5
575-225-10
1000
91.5
0.0
88.6
2.8
450-150-10
1000
91.5
0.1
91.4
1.9
300-100-10
1000
90.1
0.0
89.4
1.0
810-270-10
1000
78.1
0.0
45.6
4.4
575-225-10
1000
85.0
0.0
46.5
6.9
450-150-10
1000
87.4
0.4
50.8
5.9
300-100-10
1000
88.8
0.0
55.1
1.6
810-270-10
1000
87.5
0.0
74.1
1.3
575-225-10
1000
84.3
0.0
68.3
0.9
450-150-10
1000
83.4
0.0
71.6
0.6
300-100-10
1000
74.4
0.1
65.9
0.9
810-270-10
1000
2.1
31.9
1.6
32.8
575-225-10
1000
2.6
56.1
2.1
56.4
450-150-10
1000
3.4
8.8
2.3
9.8
300-100-10
1000
2.5
27.6
1.9
28.5
810-270-10
1000
87.5
0.0
72.3
0.3
575-225-10
1000
88.6
0.0
75.0
0.6
450-150-10
1000
91.0
0.0
80.1
0.5
300-100-10
1000
89.9
0.0
80.9
0.1
41
Ƚɪɚɮɢɤɢ ɢɡɦɟɧɟɧɢɹ ɨɲɢɛɨɤ ɨɛɭɱɟɧɢɹ ɢ ɪɚɫɩɨɡɧɚɜɚɧɢɹ ɦɨɞɟɥɢɪɭɟɦɵɯ
ɧɟɣɪɨɧɧɵɯ ɫɟɬɟɣ ɩɪɟɞɫɬɚɜɥɟɧɵ ɧɚ ɪɢɫɭɧɤɚɯ 4-7.
Ɋɢɫ. 4. Ƚɪɚɮɢɤɢ ɢɡɦɟɧɟɧɢɹ ɡɚɜɢɫɢɦɨɫɬɟɣ ɨɲɢɛɤɢ ɨɛɭɱɟɧɢɹ EL ɢ ɨɲɢɛɤɢ
ɪɚɫɩɨɡɧɚɜɚɧɢɹ EG ɩɪɢ ɢɫɩɨɥɶɡɨɜɚɧɢɢ ɚɥɝɨɪɢɬɦɚ cgb
Ɋɢɫ. 5. Ƚɪɚɮɢɤɢ ɢɡɦɟɧɟɧɢɹ ɡɚɜɢɫɢɦɨɫɬɟɣ ɨɲɢɛɤɢ ɨɛɭɱɟɧɢɹ EL ɢ ɨɲɢɛɤɢ
ɪɚɫɩɨɡɧɚɜɚɧɢɹ EG ɩɪɢ ɢɫɩɨɥɶɡɨɜɚɧɢɢ ɚɥɝɨɪɢɬɦɚ cgp
Ɋɢɫ. 6. Ƚɪɚɮɢɤɢ ɢɡɦɟɧɟɧɢɹ ɡɚɜɢɫɢɦɨɫɬɟɣ ɨɲɢɛɤɢ ɨɛɭɱɟɧɢɹ EL ɢ ɨɲɢɛɤɢ
ɪɚɫɩɨɡɧɚɜɚɧɢɹ EG ɩɪɢ ɢɫɩɨɥɶɡɨɜɚɧɢɢ ɚɥɝɨɪɢɬɦɨɜ gda
42
Ɋɢɫ. 7. Ƚɪɚɮɢɤɢ ɢɡɦɟɧɟɧɢɹ ɡɚɜɢɫɢɦɨɫɬɟɣ ɨɲɢɛɤɢ ɨɛɭɱɟɧɢɹ EL ɢ ɨɲɢɛɤɢ
ɪɚɫɩɨɡɧɚɜɚɧɢɹ EG ɩɪɢ ɢɫɩɨɥɶɡɨɜɚɧɢɢ ɚɥɝɨɪɢɬɦɨɜ gd
Ʉɚɤ ɜɢɞɧɨ, ɢɡ ɝɪɚɮɢɤɨɜ ɫ ɨɩɪɟɞɟɥɟɧɧɨɝɨ ɦɨɦɟɧɬɚ ɡɧɚɱɟɧɢɟ ɨɲɢɛɤɢ
ɪɚɫɩɨɡɧɚɜɚɧɢɹ ɧɚɱɢɧɚɟɬ ɭɜɟɥɢɱɢɜɚɬɶɫɹ, ɱɬɨ ɨɛɴɹɫɧɹɟɬɫɹ ɹɜɥɟɧɢɟɦ
ɩɟɪɟɨɛɭɱɟɧɢɹ ɧɟɣɪɨɧɧɨɣ ɫɟɬɢ [10].
Ⱦɥɹ ɩɪɟɞɨɬɜɪɚɳɟɧɢɹ ɩɪɨɰɟɫɫɚ ɩɟɪɟɨɛɭɱɟɧɢɹ ɨɛɭɱɚɸɳɟɟ ɦɧɨɠɟɫɬɜɨ
ɢɡɨɛɪɚɠɟɧɢɣ ɬɟɪɦɨɝɪɚɦɦ, ɤɚɤ ɛɵɥɨ ɨɩɢɫɚɧɨ ɜɵɲɟ, ɪɚɡɛɢɬɨ ɧɚ 2 ɦɧɨɠɟɫɬɜɚ:
ɨɛɭɱɚɸɳɟɟ ɢ ɤɨɧɬɪɨɥɶɧɨɟ.
ȼ ɪɟɡɭɥɶɬɚɬɟ ɧɚ ɨɫɧɨɜɚɧɢɢ ɝɪɚɮɢɤɨɜ ɢɡɦɟɧɟɧɢɣ ɨɲɢɛɨɤ ɨɩɪɟɞɟɥɟɧɨ
ɨɩɬɢɦɚɥɶɧɨɟ ɤɨɥɢɱɟɫɬɜɨ ɷɩɨɯ ɨɛɭɱɟɧɢɹ ɧɟɣɪɨɧɧɵɯ ɫɟɬɟɣ ɩɪɢ ɢɫɩɨɥɶɡɨɜɚɧɢɢ
ɪɚɡɧɵɯ ɚɥɝɨɪɢɬɦɨɜ ɨɛɭɱɟɧɢɹ, ɤɨɬɨɪɨɟ ɫɨɫɬɚɜɢɥɨ:
- ɞɥɹ ɚɥɝɨɪɢɬɦɚ gd - 700 ɷɩɨɯ;
- ɞɥɹ ɚɥɝɨɪɢɬɦɚ gda - 500 ɷɩɨɯ;
- ɞɥɹ ɚɥɝɨɪɢɬɦɚ cgb - 800 ɷɩɨɯ;
- ɞɥɹ ɚɥɝɨɪɢɬɦɚ cgp - 600 ɷɩɨɯ.
ɇɚ ɨɫɧɨɜɚɧɢɢ ɝɪɚɮɢɤɨɜ ɦɨɠɧɨ ɫɞɟɥɚɬɶ ɜɵɜɨɞ ɨ ɬɨɦ, ɱɬɨ ɧɚɢɛɨɥɟɟ
ɚɞɟɤɜɚɬɧɵɦ ɫɬɚɥ ɚɥɝɨɪɢɬɦ ɝɪɚɞɢɟɧɬɧɨɝɨ ɫɩɭɫɤɚ gd, ɨ ɱɟɦ ɫɜɢɞɟɬɟɥɶɫɬɜɭɟɬ
ɧɢɡɤɨɟ ɪɚɫɯɨɠɞɟɧɢɟ ɪɟɡɭɥɶɬɚɬɨɜ ɨɛɭɱɟɧɢɹ ɢ ɪɚɫɩɨɡɧɚɜɚɧɢɹ, ɱɬɨ ɩɨɞɬɜɟɪɠɞɚɟɬ
ɧɟɨɛɯɨɞɢɦɨɫɬɶ
ɢɫɩɨɥɶɡɨɜɚɧɢɹ
ɜɵɛɪɚɧɧɨɝɨ
ɚɥɝɨɪɢɬɦɚ
ɨɛɪɚɬɧɨɝɨ
ɪɚɫɩɪɨɫɬɪɚɧɟɧɢɹ ɨɲɢɛɤɢ.
ȼɵɜɨɞɵ
Ɍɚɤɢɦ ɨɛɪɚɡɨɦ, ɛɵɥɢ ɫɩɪɨɟɤɬɢɪɨɜɚɧɵ ɢ ɩɪɨɦɨɞɟɥɢɪɨɜɚɧɵ ɚɪɯɢɬɟɤɬɭɪɵ
ɦɧɨɝɨɫɥɨɣɧɵɯ ɧɟɣɪɨɧɧɵɯ ɫɟɬɟɣ, ɩɨɡɜɨɥɹɸɳɢɯ ɩɪɨɜɨɞɢɬɶ ɪɚɫɩɨɡɧɚɜɚɧɢɟ
ɢɡɨɛɪɚɠɟɧɢɣ ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɦɟɬɚɥɥɨɜ, ɜ ɱɚɫɬɧɨɫɬɢ ɫɬɚɥɟɣ ɦɚɪɨɤ 10ɏɋɇȾ,
20ɏȽɋȺ, 30ɏȽɌ, ɫ ɨɛɟɫɩɟɱɟɧɢɟɦ ɧɢɡɤɨɝɨ ɡɧɚɱɟɧɢɹ ɫɪɟɞɧɟɤɜɚɞɪɚɬɢɱɟɫɤɨɣ
ɨɲɢɛɤɢ ɪɚɫɩɨɡɧɚɜɚɧɢɹ. ɇɚɢɛɨɥɶɲɭɸ ɷɮɮɟɤɬɢɜɧɨɫɬɶ ɜ ɩɪɨɜɟɞɟɧɧɵɯ
ɷɤɫɩɟɪɢɦɟɧɬɚɯ ɪɚɫɩɨɡɧɚɜɚɧɢɹ ɩɨɤɚɡɚɥɢ ɦɧɨɝɨɫɥɨɣɧɵɟ ɫɟɬɢ, ɨɛɭɱɟɧɧɵɟ ɩɨ
ɚɥɝɨɪɢɬɦɭ ɝɪɚɞɢɟɧɬɧɨɝɨ ɫɩɭɫɤɚ gd.
43
1. Bramfitt B. L. Metallographer's Guide – Practices and Procedures for Iron and Steels / B.
L. Bramfitt, Arlan O. Benscoter. – ASM International, 2002. – 354 p.
2. Hosseini H. Characterization of microstructures and mechanical properties of Inconel
617/310 stainless steel dissimilar welds./ H. Hosseini, M. Shamaniana, A. Kermanpura //
Materials Characterization. – Vol. 62, Issue 4. – 2011. – P. 425 – 431.
3. Cast-iron metallographic structure by computer picture processing system / Wang
Zhiping, Lu Yang, Wu Chenwed, Xu Jianlin, Yang Xinzhuang // Journal of Cansu University
of Technology, – Vol.E-1, No.1. – 1997. – p. 29 – 32.
4. VideoTesT Ltd “Application of Image analysis software “VideoTesT-Morphology” in
mycology, phytopathology and industrial microbiology” [ɗɥɟɤɬɪɨɧɧɵɣ ɪɟɫɭɪɫ] – Ɋɟɠɢɦ
ɞɨɫɬɭɩɚ: http://www.videotest.ru/en/article/ view/48/category/11 (26.08.2007)
5. VisionPE Metlab Corporation. Image Analysis, [ɗɥɟɤɬɪɨɧɧɵɣ ɪɟɫɭɪɫ] – Ɋɟɠɢɦ
ɞɨɫɬɭɩɚ: http://www.metlabcorp.com/image_analysis.html (31.01.2009)
6. SIAMS: Index of/products/siams700 [ɗɥɟɤɬɪɨɧɧɵɣ ɪɟɫɭɪɫ] – Ɋɟɠɢɦ ɞɨɫɬɭɩɚ: http://
www.siams.com/products/siams700/ (14.01.2013)
7. ɀɭɪɚɜɥɟɜ ɘ. ɂ. Ɋɚɫɩɨɡɧɚɜɚɧɢɟ. Ɇɚɬɟɦɚɬɢɱɟɫɤɢɟ ɦɟɬɨɞɵ. ɉɪɨɝɪɚɦɦɧɚɹ ɫɢɫɬɟɦɚ.
ɉɪɚɤɬɢɱɟɫɤɢɟ ɩɪɢɦɟɧɟɧɢɹ / ɘ. ɂ. ɀɭɪɚɜɥɟɜ, ȼ. ȼ. Ɋɹɡɚɧɨɜ, Ɉ. ȼ. ɋɟɧɶɤɨ. – Ɇ. :
Ɏɚɡɢɫ, – 2005. – 159 ɫ.
8. Gonzalez R.S. Digital image processing. – 2nd ed. / R.S. Gonzalez, R.E. Woods. – USA:
Prentice, 2002. – 703 p.
9. ȿɦɟɥɶɹɧɨɜ ȼ. Ⱥ. ɇɟɣɪɨɫɟɬɟɜɨɣ ɦɟɬɨɞ ɨɩɪɟɞɟɥɟɧɢɹ ɤɨɥɢɱɟɫɬɜɟɧɧɵɯ ɯɚɪɚɤɬɟɪɢɫɬɢɤ
ɦɟɬɚɥɥɨɜ / ȼ.Ⱥ. ȿɦɟɥɶɹɧɨɜ // Ɋɚɞɿɨɟɥɟɤɬɪɨɧɧɿ ɿ ɤɨɦɩ’ɸɬɟɪɧɿ ɫɢɫɬɟɦɢ. – 2010. – ʋ
4(45). – ɋ.169 - 173.
10. Suzuki Kenji Artificial Neural Networks: Architectures and Applications [Text] / Kenji
Suzuki. – InTech, 2013. – 256 p.
11. ɋɚɣɬ: Ȼɚɡɚ ɞɚɧɧɵɯ ɦɢɤɪɨɫɬɪɭɤɬɭɪ ɦɟɬɚɥɥɨɜ ɢ ɫɩɥɚɜɨɜ. [ɗɥɟɤɬɪɨɧɧɵɣ ɪɟɫɭɪɫ] –
Ɋɟɠɢɦ ɞɨɫɬɭɩɚ: http://www.microstructure.ru/rudbview
ɉɨɫɬɭɩɢɥɚ 3.02.2014ɪ.
ɍȾɄ 004.382.76
Ⱦ.ȼ. ɋɬɚɫɶ, Ʉɢʀɜ
ɉɊɈȽɊȺɆɇȱ ɁȺɋɈȻɂ ɈɐȱɇɄɂ ȿɇȿɊȽɈɁȻȿɊȿɀȿɇɇə ɆɈȻȱɅɖɇɂɏ
ɉɊɂɋɌɊɈȲȼ
Abstract. In this paper, the theoretical information about mobile device energy
consumption was presented. The methods of energy saving implemented in modern
mobile operating systems were reviewed. Programmatic methods for measuring
energy consumption of mobile devices were investigated and compared based on iOS
and Android mobile operating systems.
ɉɨɫɬɿɣɧɨ ɡɪɨɫɬɚɸɱɿ ɦɨɠɥɢɜɨɫɬɿ ɫɦɚɪɬɮɨɧɿɜ, ɩɥɚɧɲɟɬɿɜ, ɤɨɦɩɚɤɬɧɢɯ ɉɄ
ɜɢɦɚɝɚɸɬɶ ɜɿɞ ɜɢɪɨɛɧɢɤɿɜ ɦɨɛɿɥɶɧɢɯ ɩɪɢɫɬɪɨʀɜ ɜɢɤɨɪɢɫɬɚɧɧɹ ɫɩɟɰɿɚɥɶɧɢɯ
44 © Ⱦ.ȼ. ɋɬɚɫɶ
1/--страниц
Пожаловаться на содержимое документа