Minyak sawit dihasilkan di kilang kelapa sawit yang dilengkapi loji kuasa stim tersendiri dan loji terbabit menggunakan bahan buangan kelapa sawit (sabut dan tempurung) sebagai bahan api dandang. Walau bagaimanapun, hasil pembakaran menyebabkan pencemaran ke atmosfera yang serius. Pelepasan asap melalui serobong boleh dipantau dengan menyelaku proses masukan (dalam bahan api, turbin, dandang) dan keluaran pencemar. Dalam kertas kerja ini, Rangkaian Neural Buatan (ANN) digunakan untuk menyelaku asap dari dandang kilang kelapa sawit. Regresi Lelurus Pelbagai (MLR) juga digunakan untuk mencari pekali unsur yang menyumbang kepada pelepasan setiap pencemar dan membanding dan mengesahkan keputusan ANN. Kesimpulannya, ramalan yang dibuat oleh ANN lebih baik daripada MLR tetapi keduanya menunjukkan keputusan yang hampir sama dengan nilai sebenar yang diperolehi dari kilang. Kata kunci: Rangkaian neural buatan; emisi dandang biojisim; regrasi lelurus pelbagai; ramalan dan perbandingan Palm oil is produced in palm oil mills, where palm oil waste can be used (shell and fibre) as fuel for the boilers for generating steam power plants. Unfortunately, the combustion products of these materials cause severe atmospheric pollutions. The emission released through the chimney can be monitored by modeling its process of input (in fuel, turbine, boiler) and output of the pollutants. In this paper, Artificial Neural Networks (ANN) is used to model the emission from the palm oil mill boiler. Multiple Linear Regression (MLR) is also applied to find the coefficient of the contributing element to the pollution in order to make comparison and validate the ANN results. In conclusion, the prediction made by ANN is better than MLR but both agrees well with the actual values collected from the mill. Key words: Artificial neural network; biomass boiler emission; multiple linear regression; prediction and comparison
Purpose This study aims to investigate the interaction of independent variables [Reynolds number (Re), thermal power and the number of ball grid array (BGA) packages] and the relation of the variables with the responses [Nusselt number ((Nu) ¯ ), deflection/FPCB’s length (d/L) and von Mises stress]. The airflow and thermal effects were considered for optimizing the Re of various numbers of BGA packages with thermal power attached on flexible printed circuit board (FPCB) for optimum cooling performance with least deflection and stress by using the response surface method (RSM). Design/methodology/approach Flow and thermal effects on FPCB with heat source generated in the BGA packages have been examined in the simulation. The interactive relationship between factors (i.e. Re, thermal power and number of BGA packages) and responses (i.e. deflection over FPCB length ratio, stress and average Nusselt number) were analysed using analysis of variance. RSM was used to optimize the Re for the different number of BGA packages attached to the FPCB. Findings It is important to understand the behaviour of FPCB when exposed to both flow and thermal effects simultaneously under the operating conditions. Maximum d/L and von Misses stress were significantly affected by all parametric factors whilst (Nu)¯ is significantly affected by Re and thermal power. Optimized Re for 1–3 BGA packages with maximum thermal power applied has been identified as 21,364, 23,858 and 29,367, respectively. Practical implications This analysis offers a better interpretation of the parameter control in FPCB with optimized Re for the use of force convection electronic cooling. Optimal Re could be used as a reference in the thermal management aspect in designing the BGA package. Originality/value This research presents the parameters’ effects on the reliability and heat transfer in FPCB design. It also presents a method to optimize Re for the different number of BGA packages attached to increase the reliability in FPCB’s design.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.