CAN liidesega aku monitoorimis- ja andmelogimissüsteem põllumajandusrobotitele
Laen...
Kuupäev
2025
Kättesaadav alates
13.09.2025
Autorid
Ajakirja pealkiri
Ajakirja ISSN
Köite pealkiri
Kirjastaja
Eesti Maaülikool
Abstrakt
Käesolevas artiklis tutvustatakse kerge ja energiatõhusa akujälgimise ja logimise süsteemi kavandamist, arendust ja hindamist, mis on spetsiaalselt kohandatud põllumajandusrobotite vajadustele. Süsteem on plaanitud töötama võrguühenduseta välitingimustes ning selle modulaarne arhitektuur sisaldab CAN-võrgu (Controller Area Network) põhist aku haldussüsteemi (BMS) ühendust, Raspberry Pi mikrokontrollerit, madala energiatarbega e–ink ekraani kohapealseks visuaalseks tagasisideks ning veebipõhist API-liidest kaugdiagnostikaks. Aku põhitähtsusega tööparameetrid – pinge, voolutugevus, temperatuur ja laetuse tase (SoC) – kogutakse CAN-võrgu kaudu ning salvestatakse SQLite andmebaasi. Süsteem on täielikult konfigureeritav ja ühilduv nii CAN-põhiste BMS-seadmetega kui ka avatud platvormidega, nagu Robot Operating System (ROS). Süsteemi keskne uuendus seisneb adaptiivses andmekogumise algoritmis, mis reguleerib andmete pärimise sagedust vastavalt aku töörežiimile ja temperatuuri tasemetele. See võimaldab vähendada energiatarbimist ilma süsteemi reageerimisvõimet ohverdamata. Lisaks reaalajas jälgimisele seisneb süsteemi pikaajaline väärtus struktureeritud andmestikus, mis võimaldab tulevikus rakendada tehisintellektil põhinevaid diagnostikameetodeid, ennustavat hooldust ja dünaamilist juhtimisloogikat. Tööläved ja uuendussagedused on kasutaja poolt andmebaasi kaudu seadistatavad, võimaldades süsteemi täpset kohandamist välitingimustele. Esialgsed katsed kinnitavad süsteemi võimet tuvastada kõrvalekaldeid, toetada ajaloolist diagnostikat ja vähendada energiakulu läbi intelligentse ajastuse. Tänu avatud ja riistvarast sõltumatule ülesehitusele on platvormi võimalik hõlpsasti rakendada erinevates CAN-ühilduvates süsteemides.
This article presents the design, implementation, and evaluation of a lightweight, energy-efficient battery monitoring and logging system tailored for agricultural robotics. Targeting off-grid field operations, the system features a modular architecture integrating a Controller Area Network (CAN) - based Battery Management System (BMS), a Raspberry Pi microcontroller, a low-power e-ink display for local feedback, and a web-accessible API for remote diagnostics. Core battery parameters - voltage, current, temperature, and state of charge - are collected via the CAN bus and logged to an onboard SQLite database. The system is fully configurable, lightweight, and modular, supporting compatibility with CAN-based BMS devices and open platforms like Robot Operating System (ROS). A key innovation is its adaptive data acquisition algorithm, which adjusts polling frequency based on battery activity and temperature thresholds, significantly reducing power consumption without compromising responsiveness. Beyond real-time monitoring, the system’s primary value lies in the structured dataset it generates. This long-term data enables future applications such as AI-based diagnostics, predictive maintenance, and adaptive control strategies. All operational thresholds and refresh rates are user-adjustable via the database, allowing precise tuning to field conditions. Preliminary tests confirm the system’s ability to detect anomalies, support historical diagnostics, and reduce energy consumption through intelligent scheduling. The hardware-agnostic and non-proprietary approach makes the platform scalable and adaptable to a wide range of CAN-compatible systems. By combining modular design, dynamic data logging, and remote access, the system advances sustainable battery management in agricultural robotics and creates a foundation for future integration with autonomous systems and machine learning models.
This article presents the design, implementation, and evaluation of a lightweight, energy-efficient battery monitoring and logging system tailored for agricultural robotics. Targeting off-grid field operations, the system features a modular architecture integrating a Controller Area Network (CAN) - based Battery Management System (BMS), a Raspberry Pi microcontroller, a low-power e-ink display for local feedback, and a web-accessible API for remote diagnostics. Core battery parameters - voltage, current, temperature, and state of charge - are collected via the CAN bus and logged to an onboard SQLite database. The system is fully configurable, lightweight, and modular, supporting compatibility with CAN-based BMS devices and open platforms like Robot Operating System (ROS). A key innovation is its adaptive data acquisition algorithm, which adjusts polling frequency based on battery activity and temperature thresholds, significantly reducing power consumption without compromising responsiveness. Beyond real-time monitoring, the system’s primary value lies in the structured dataset it generates. This long-term data enables future applications such as AI-based diagnostics, predictive maintenance, and adaptive control strategies. All operational thresholds and refresh rates are user-adjustable via the database, allowing precise tuning to field conditions. Preliminary tests confirm the system’s ability to detect anomalies, support historical diagnostics, and reduce energy consumption through intelligent scheduling. The hardware-agnostic and non-proprietary approach makes the platform scalable and adaptable to a wide range of CAN-compatible systems. By combining modular design, dynamic data logging, and remote access, the system advances sustainable battery management in agricultural robotics and creates a foundation for future integration with autonomous systems and machine learning models.
Kirjeldus
Magistritöö
Energiakasutuse õppekaval
Märksõnad
magistritööd, põllumajandusrobootika, akuhaldussüsteem (BMS), CAN-siini jälgimine, andmelogimine, manustatud süsteemid, energiatõhusus