Zeta Binta Mohammed Adib, Alam Sifat U, Rahman Gazi M A Ehsan Ur, Ahmed Khawza Iftekhar Uddin
Department of Electrical and Electronic Engineering, United International University, Dhaka 1212, Bangladesh.
Department of Electrical and Electronic Engineering, Green University of Bangladesh, Narayanganj 1461, Bangladesh.
Sensors (Basel). 2025 Apr 30;25(9):2824. doi: 10.3390/s25092824.
For aquaculture systems, pH is the prime quality indicator and is highly related to other water quality indicators like ammonia and ammonium ions. The available pH sensors using chemical references are not suitable for continuous in situ monitoring of aquaculture systems due to their frequent calibration requirement and high cost. This research develops a pH sensor with temperature compensation implementing a machine learning (ML) algorithm. Unlike traditional methods, this sensor utilizes electronic calibration, eliminating the need for chemical calibration and ongoing maintenance efforts. The application of this low-cost sensor is particularly well suited for in situ aquaculture scenarios, where multiple local sensor nodes operate under the control of a master node. The test results of the developed sensor show an improved sensitivity from 0.288 µA/pH to 0.316 µA/pH compared to the available pH sensors. Additionally, the response time has been improved from 1 s to 125 ms, showcasing the suitability of this pH sensor for real-time water quality monitoring of aquaculture applications.
对于水产养殖系统而言，pH值是首要的水质指标，且与氨和铵离子等其他水质指标高度相关。现有的使用化学参比的pH传感器由于频繁校准的要求和高昂的成本，不适用于水产养殖系统的连续原位监测。本研究开发了一种采用机器学习（ML）算法实现温度补偿的pH传感器。与传统方法不同，该传感器采用电子校准，无需化学校准和持续维护工作。这种低成本传感器的应用特别适合原位水产养殖场景，其中多个本地传感器节点在主节点的控制下运行。所开发传感器的测试结果表明，与现有pH传感器相比，灵敏度从0.288 µA/pH提高到了0.316 µA/pH。此外，响应时间从1秒缩短到了125毫秒，表明这种pH传感器适用于水产养殖应用的实时水质监测。
