BIUSTRE

Innovative hydroponic control using hybrid intelligent algorithms: a case study in Botswana

Show simple item record

dc.contributor.supervisor Matsebe, O.
dc.contributor.supervisor Ditshego, N.
dc.contributor.author Rathedi, Maemo
dc.date.accessioned 2026-03-24T08:34:50Z
dc.date.available 2026-03-24T08:34:50Z
dc.date.issued 2025-04
dc.identifier.citation Rathedi,M.(2025) Innovative hydroponic control using hybrid intelligent algorithms: a case study in Botswana, Master’s thesis, Botswana International University of Science and Technology: Palapye en_US
dc.identifier.uri https://repository.biust.ac.bw/handle/123456789/738
dc.description.abstract This work evaluates various control algorithms in hydroponics farming, to analyse the performance of the algorithms with the aim to design an intelligent controller which can improve resource efficiency and crop yield by leveraging their ability to handle non linear systems more effectively than conventional controllers. The focus is on the nutrient solution section of the hydroponics farm, specifically monitoring and controlling pH, water level, and water temperature. To achieve this, both simulation and prototyping approaches are adopted. First, mathematical models for each environmental parameter are developed using first-order differential equations in the MATLAB Simulink platform. Conventional PID control, as well as intelligent controllers such as PID-Fuzzy Logic Control (PID-FLC), Fuzzy Logic Control (FLC), are designed and applied to the developed transfer functions. The performance of these algorithms are then evaluated using transient performance parameters: steady- state error, overshoot, settling time, and rise time. These metrics are used to compare the control systems against practical requirements set by farmers or users. Based on a systematic approach, the best-performing algorithm is selected for further evaluation in a prototype system together with a data driven control algorithm . The prototype consists of a 24-plant nutrient film technique (NFT) hydroponic system. Data collected from this setup are used to implement a data-driven Long Short Term Memory-Artificial Neural Network (LSTM-ANN) controller, which is then compared against the selected control algorithm from the simulation stage. The research showed that PID-FLC performed better in terms of responsiveness recording lower rise times and setting times. Specifically, the PID-FLC recorde rise times of 60 seconds for pH increase and 70 seconds for pH decrease, compared to the ANN’s 230 seconds and 120 seconds, respectively. For settling times pH up for PID-FLC 119.38 seconds ANN 320 seconds. In terms of steady-state error, all algorithms showed capabilities in maintaining required levels as they recorded low values close to 0 or range close to the setpoint. As for overshoot, the controllers were able to maintain overshoots of less than 2.5%. The study further analyses the financial viability of the use of an intelligent controller analysing the cost savings, net present value, and payback period of the project. en_US
dc.language.iso en en_US
dc.publisher Botswana International University of Science and Technology (BIUST) en_US
dc.subject Hydroponic control en_US
dc.subject Hybrid intelligent algorithms en_US
dc.subject Hydroponics farming en_US
dc.subject Intelligent control systems en_US
dc.subject PID-Fuzzy Logic Control (PID-FLC) en_US
dc.subject Fuzzy Logic Control (FLC) en_US
dc.subject Long Short-Term Memory Artificial Neural Network (LSTM-ANN) en_US
dc.subject Nutrient solution monitoring and control en_US
dc.title Innovative hydroponic control using hybrid intelligent algorithms: a case study in Botswana en_US
dc.description.level phd en_US
dc.description.accessibility unrestricted en_US
dc.description.department mie en_US


Files in this item

Icon
Name: Rathedi_MEng_2025.pdf
Size: 7.003Mb
Format: PDF
View/Open

This item appears in the following Collection(s)

Show simple item record

Search BIUSTRE


Browse

My Account