CD Tesis

Prediksi Tingkat Kinerja Struktur Bangunan Berdasarkan Mutu Beton Dengan Metode Jaringan Saraf Tiruan

---

Most conditions of multi-story buildings in Bangkinang City are rated not meet the requirements of the planning standards because they are carried out without including earthquake loads during structural planning. This study was conducted aimed to analyze the performance level of building structures based on reviews of displacement, velocity, and acceleration of earthquake loads, as well as identifying the accuracy of Artificial Neural Network (ANN) method in predicting the level of damage to buildings using time history analysis. The object of research is one of the Office Buildings in Bangkinang City. The analysis used is the Non Linear Time History (NLTH) to obtain the performance level of the building structure, which in the planning stage does not calculate the earthquake load, then the analysis process is carried out by modeling the building in a variety of concrete quality between fc'15 MPa to 25 MPa, with the Elcentro earthquake scaling 0.25g, 0.5g, 0.75g, 1g and the peak land acceleration of Bangkinang City 0.024g. Furthermore, the results of the analysis of the building's time history will be analyzed using the Artificial Neural Network (ANN) method, with the result is a level of damage to buildings. After predicting with the Artificial Neural Network (ANN) method, the prediction results are validated using the Mean-Squared Errors (MSE) and Determination Coefficients (R2). The response of the building structure to the Bangkinang peak ground acceleration scale (0.024g) and with all variations in the strenght of concrete, when the THNL analysis was carried out the structure was not damaged and did not reach the condition where the damage began to occur, which means the structural conditions are very safe. Whereas for a 0.25g scale for all concrete quality variations, the structure starts to experience minor damage but the structure is still very suitable for use. The results of the analysis with 0.5g, 0.75g and 1g earthquake scale for all concrete quality were immediately destroyed after the damage began to occur without experiencing Life Safety or Collapse Prevention symptoms. The Artificial Neural Network (ANN) in predicting building structure performance with accuracy (R2) ranges from 94,952% to 98,119%, and the Mean-Squared Errors (MSE) value is 0,00025 for training data sets and 0,00082 for testing data sets . According to to the result the ANN method is very capable of predicting the response of building structures that are very well reviewed.
Keywords: Performance structure, time history analysis, artificial neural network

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

10 12. 118 (0007)

Penerbit

Pekanbaru : Universitas Riau – Pascasarjana – Tesis Teknik Sipil., 2018

Deskripsi Fisik

xiv, 68 hlm.; ill.; 29 cm

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

10 12. 118 (0007)

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

-

Subjek

PASCASARJANA (MAGISTER) TEKNIK SIPIL

Info Detail Spesifik

-

Pernyataan Tanggungjawab

FATAH

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar