CD Skripsi

Prediksi Kadar Particulate Matter (Pm10)Untuk Pemantauan Kualitas Udara Menggunakan Jaringan Syaraf Tiruan

---

Air pollution in Pekanbaru City, specifically at PM10 levels, occurs every year. A PM10 level prediction analysis was conducted using Artificial Neural Network backpropagation based on weather parameters. The goal is to monitor air quality as an anticipatory step. The data used was collected based on monthly data from the year of 2014 to 2018, including PM10 data, precipitation, wind speed, air temperature, humidity and sun exposure period. This method used R2015a matlab software for data programming, Microsoft Excel is used for data grouping and sharing, and Sigmaplot is utilized for data graphics. The architecture used consists of 5 input layers using the Log Activation function, 5 hidden layers using the Log Activation function, and 1 transformation layer using Purelin function. This analysis divides the data into two sections, namely data from 2014 to 2017 as training data and data from 2018 as test data. The results of the network training study, namely the CGB train, provided MSE values of -0.0705 and the best PM10 estimates for the network testing in February, with an error percentage of -2.0526 per cent with a broad PM10 of 44.222µm/m3 and a PM10 BMKG data of 45.136µm/m3. Although the forecast results for artificial neural networks with the biggest error namely 91.752 per cent in November. The average forecast error for artificial neural networks for 1 year was 26.9062 per cent. The magnitude of the error in November due to the high rainfall during that month, so that it has reduced the PM10 BMKG level.

Keywords : Particulate Matter (PM10), Artificial Neural Network, Back Propagation, Prediction.

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

03 01.120 (0019)

Penerbit

Pekanbaru : Universitas Riau – FMIPA – FISIKA., 2020

Deskripsi Fisik

xv, 48 hlm.: ill.: 29 cm

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

03 01.120 (0019)

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

-

Subjek

FISIKA

Info Detail Spesifik

-

Pernyataan Tanggungjawab

ELFITRA

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar