Solving sensor instability in an automatic dosing system for vertical farming

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu | Master's thesis
Date
2024-05-20
Department
Major/Subject
Control, Robotics and Autonomous Systems
Mcode
ELEC3025
Degree programme
AEE - Master’s Programme in Automation and Electrical Engineering (TS2013)
Language
en
Pages
47
Series
Abstract
Vertical farming has been trending in recent years and it has been offered as a solution to growing food locally in urban areas. In the most basic terms, this means growing food on multiple levels in an indoor environment which means that the farming requires solutions in both irrigation and lighting. While artificial lighting has become relatively cheap and mainstream after LEDs entered the markets, artificial irrigation indoors on multiple levels is not as straightforward. One solution to this is hydroponic irrigation methods. Hydroponic irrigation has been researched and developed which means that the plants can be grown without soil. The lack of soil however means that nutrients need to be added to the irrigation water. Maintaining the correct levels of nutrients in irrigation water is a repetitive task and easily prone to human error, so an automatic dosing system has been developed to take care of it. The automatic dosing system needs to know the current pH level of the water but the available pH probes are not accurate enough for this application and their values fluctuate a lot. This thesis aims to filter the pH values in a way that they can reliably be used in an automatic dosing system in vertical farming. In total four methods were tested to stabilize the pH values. These methods were the moving average (MA) method, ūĚõľ-ūĚõĹ-filter, ūĚõľ-ūĚõĹ-filter with an implemented threshold, and the Kalman filter (KF). Each method was tested using two sensors from different manufacturers both in a stable environment and in a changing one. Out of these methods, the Kalman filter proved to be the best fit for our system.

Vertikaaliviljely on kasvattanut suosiotaan viime aikoina ja sit√§ usein tarjotaan ratkaisuna ruoantuotantoon paikallisesti sis√§tiloissa kaupunkialueilla. Yksinkertaistettuna vertikaaliviljely tarkoittaa ruuan kasvattamista monessa kerroksessa sis√§tiloissa kontrolloidussa ymp√§rist√∂ss√§, mik√§ vaatii ratkaisuja niin valaistuksen kuin kastelunkin suhteen. Vaikka keinotekoinen valaistus onkin nyky√§√§n suhteellisen edullista ja helposti saatavilla LEDien yleistytty√§ markkinoilla, keinotekoinen kastelu useassa kerroksessa ei ole viel√§ samalla tasolla. Yksi ratkaisu t√§h√§n on hydroponiset kastelyj√§rjestelm√§t. Hydroponisia kastelumenetelmi√§ on tutkittu ja kehitetty paljon siihen pisteeseen, ett√§ kasveja voidaan helposti kasvattaa ilman multaa. Mullan puute kuitenkin tarkoittaa, ett√§ kasteluveteen tulee lis√§t√§ ravinteita. Oikean ravinnetason yll√§pit√§minen kasteluvedess√§ on yksitoikkoinen teht√§v√§, jossa on helppo tehd√§ virheit√§. T√§m√§n vuoksi sen hoitamiseksi on kehitetty automaattinen sy√∂tt√∂j√§rjestelm√§. Sy√∂tt√∂j√§rjestelm√§n tulee tiet√§√§ kasteluveden pH-taso, mutta t√§ll√§ hetkell√§ markkinoilta saatavat pH-sensorit eiv√§t ole tarpeeksi tarkkoja t√§h√§n teht√§v√§√§n. T√§m√§n diplomity√∂n tarkoituksena on suodattaa pH-arvoja siten, ett√§ niit√§ voidaan k√§ytt√§√§ luotettavasti sy√∂tt√∂j√§rjestelm√§ss√§. Yhteens√§ nelj√§√§ eri metodia kokeiltiin pH-arvojen tasaamiseksi. N√§m√§ metodit olivat liikkuva keskiarvo (Moving Average), ūĚõľ-ūĚõĹ-suodatin, ūĚõľ-ūĚõĹ-suodatin kynnyksell√§ ja Kalman suodatin (Kalman filter). Jokaista metodia testattiin kahden eri valmistajan pH-sensorien avulla sek√§ stabiilissa ymp√§rist√∂ss√§ ett√§ muuttuvassa ymp√§rist√∂ss√§. Kaikista testatuista metodeista Kalman suodatin sopi parhaiten systeemiimme.
Description
Supervisor
Baumann, Dominik
Thesis advisor
Romanov-Chernigovsky, Ignaty
Keywords
vertical farming, pH probe, Kalman filter, alpha-beta-filter, moving average, sensor instability
Other note
Citation