FAUZIAH, Farah and Surarso, Bayu and Tarno, Tarno (2026) ADAPTIVE NEURO FUZZY INFERENCE SYSTEM (ANFIS) DENGAN OPTIMASI FIREFLY ALGORITHM (FA) DAN PARTICLE SWARM OPTIMIZATION (PSO) UNTUK PREDIKSI DATA INDEKS KUALITAS UDARA. Masters thesis, UNIVERSITAS DIPONEGORO.

	Text cover aja.pdf Download (152kB)
	Text cOVER1.pdf Restricted to Repository staff only Download (1MB)
	Text BAB I.pdf Download (270kB)
	Text BAB II.pdf Download (563kB)
	Text BAB III.pdf Restricted to Repository staff only Download (585kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (1MB)
	Text BAB V.pdf Restricted to Repository staff only Download (216kB)
	Text Dafar pustaka.pdf Download (310kB)
	Text Lampiran.pdf Restricted to Repository staff only Download (1MB)

Abstract

Kualitas udara merupakan salah satu faktor penting yang berpengaruh terhadap kesehatan manusia, kesejahteraan masyarakat, serta keberlanjutan lingkungan. Peningkatan emisi yang berasal dari aktivitas industri, transportasi, dan perubahan penggunaan lahan telah menyebabkan penurunan kualitas udara di berbagai wilayah. Kondisi ini mendorong perlunya metode prediksi yang andal untuk memperkirakan konsentrasi polutan udara sehingga dapat mendukung upaya mitigasi serta pengelolaan lingkungan secara lebih efektif. Penelitian ini bertujuan untuk mengembangkan model prediksi kualitas udara menggunakan metode Adaptive Neuro Fuzzy Infrence System (ANFIS) yang dioptimasi dengan algoritma Particle Swarm Optimization (PSO) dan Firefly Algorithm (FA). Pendekatan penelitian dilakukan secara eksperimental melalui pemodelan data deret waktu (time series) dengan memanfaatkan empat parameter polutan utama, yaitu Karbon Monoksida (CO), Nitrogen Dioksida (NO2), Sulfur Dioksida (SO2) dan Ozon (O3). Data diolah melalui tahapan praproses, pembentukan model ANFIS serta penerapan algoritma optimasi untuk meningkatkan aurasi parameter model. Evaluasi kinerja model dilakukan menggunakan beberapa metrik, yaitu MSE, RMSE, R2 serta metrik kesalahan relatif MAPE dan SMAPE. Hasil penelitian menunjukkan bahwa model hibrida ANFIS – PSO – FA menghasilkan performa prediksi terbaik dibandingkan model lainnya. Model tersebut memperoleh nilai MSE sebesar 0,0010 RMSE sebesar 0,0320 dan R2 sebesar 0,9767 yang menunjukkan tingkat akurasi prediksi yang sangat baik. Integrasi algoritma optimasi dalam ANFIS terbukti mampu meningkatkan kemampuan model dalam mempelajari pola nonlinier pada data kualitas udara. Model ANFIS – PSO – FA berpotensu digunakan sebagai alat bantu dalam sistem pemantauan kualitas udara serta mendukung pengambilan keputusan dalam pengendalian pencemaran udara dan perlindungan kesehatan masyarakat.
Kata Kunci : Kualitas udara, ANFIS, Algoritma Firefly, Particle Swarm Optimization, Akurasi Prediksi

Air quality is one of the important factors that affect human health, community welfare, and environmental sustainability. Increased emissions from industrial activities, transportation, and land-use change have led to a decline in air quality in various regions. This condition encourages the need for reliable prediction methods to estimate the concentration of air pollutants so that it can support environmental mitigation and management efforts more effectively. This study aims to develop an air quality prediction model using the Adaptive Neuro Fuzzy Infrence System (ANFIS) method which is optimized with the Particle Swarm Optimization (PSO) and Firefly Algorithm (FA) algorithms. The research approach was carried out experimentally through time series data modeling by utilizing four main pollutant parameters, namely Carbon Monoxide (CO), Nitrogen Dioxide (NO2), Sulfur Dioxide (SO2) and Ozone (O3). Data is processed through the preprocessing stage, ANFIS model formation and the application of optimization algorithms to increase the auras of model parameters. The evaluation of model performance was carried out using several metrics, namely MSE, RMSE, R2 as well as the relative error metrics of MAPE and SMAPE. The results show that the ANFIS – PSO – FA hybrid model produces the best prediction performance compared to other models. The model obtained an MSE value of 0.0010, an RMSE of 0.0320 and an R2 of 0.9767 which indicates an excellent level of prediction accuracy. The integration of optimization algorithms in ANFIS has been proven to improve the model's ability to study nonlinear patterns in air quality data. The ANFIS – PSO – FA model has the potential to be used as an aid in air quality monitoring systems and support decision-making in air pollution control and public health protection.
Keywords : Air quality, ANFIS, Firefly Algorithm, Particle Swarm Optimization, Prediction Accuracy

Actions (login required)

View Item