Advancements in fuzzy expert systems for site-specific nitrogen fertilisation: Incorporating RGB colour codes and irrigation schedules for precision maize production in Bangladesh

Autor: Bitopi Biswas, Mohammad Tariful Alam Khan, Mohammad Billal Hossain Momen, Mohammad. Rashedur Rahman Tanvir, Abu Mohammad Shahidul Alam, M Robiul Islam Islam
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: Research in Agricultural Engineering, Vol 70, Iss 3, Pp 155-166 (2024)
Druh dokumentu: article
ISSN: 1212-9151
1805-9376
DOI: 10.17221/35/2024-RAE
Popis: The research was conducted at the Department of Agronomy and Agricultural Extension, Rajshahi University, from December 2021 to April 2022. The objective was to develop a fuzzy expert system for site-specific N fertilisation using leaf colour code (RGB) and irrigation frequencies for maize yield. The experiment encompassed two primary factors: nitrogen fertiliser application rates (N1: 100%, N2: 75%, N3: 50% of conventional rates) and irrigation frequencies (I1: 100%, I2: 75%, I3: 50% of pan evaporation). A completely randomised design (CRD) with three replications was used to arrange the experimental pots, each receiving recommended doses of phosphorus, potassium, and sulfur, with urea applied per treatment instructions. Results revealed significant chlorophyll content and grain yield differences among the various nitrogen fertiliser rates. The highest grain yield (219.27 g.pot-1) was observed with N1, whereas the lowest (186.6 g.pot-1) was with N3. Similarly, irrigation frequencies significantly influenced chlorophyll content and cob characteristics, with I1 resulting in the highest grain yield (211.27 g.pot-1) and I3 the lowest (184.6 g.pot-1). Furthermore, the interaction between fertiliser application rates and irrigation frequencies had notable effects on various parameters, leading to the highest grain yield of 227.62 g.pot-1 with the combination of N1 and I1 and the lowest (168.00 g.pot-1) with N3 I3. The agricultural experiments were facilitated using the Matlab fuzzy toolbox, employing the Mamdani inference method. Fuzzy rules were delineated for nitrogen application rates and irrigation frequencies, with three fuzzy sets each. Membership functions were developed utilising Matlab's fuzzy interface system (FIS) editor and membership function editor, optimising leaf chlorophyll content, evaporation rate as input tiger N fertilisation, and irrigation frequencies as output for precise maize production in Bangladesh.
Databáze: Directory of Open Access Journals
Externí odkaz: