Journal of Instrumentation and Innovation Sciences

Performance Comparison of Kalman Filter and Extended Kalman Filter for Sensor Fusion

2026-06-20T12:04:28+00:00

Embedded systems depend on sensor fusion because it enhances their ability to estimate states accurately and reliably through multiple sensor measurements, which contain errors. The research paper establishes performance benchmarks for the Kalman Filter (KF) and the Extended Kalman Filter (EKF) as they operate within embedded sensor fusion systems. The KF establishes an optimal method for recursive estimation of linear systems that experience Gaussian noise, while the EKF enables nonlinear system estimation through its first-order linearization method, which uses Taylor series expansion. The unified framework, which supports both filters on embedded platforms, addresses three key constraints, namely limited computational resources, limited memory, and real-time processing requirements. The research study uses multi-sensor data simulation to assess three factors, which include estimation accuracy, convergence rate and computational complexity. The results demonstrate that KF operates efficiently with low computational requirements in linear systems, while EKF delivers better accuracy for nonlinear situations but needs more complex resources to function. The study results demonstrate that there exists an essential relationship between accuracy and processing efficiency. The KF method works best for systems that have restricted computational power and show behavior that closely resembles linearity, while the EKF method becomes necessary for handling nonlinear system behavior.

Performance Analysis of Real-Time Solar Data Acquisition Systems

2026-07-08T06:53:40+00:00

The increasing deployment of photovoltaic (PV) systems has created a growing demand for intelligent monitoring solutions capable of providing accurate and continuous performance assessment. This paper presents a comprehensive performance analysis of a Real-Time Solar Data Acquisition System (RTSDAS) designed for PV energy monitoring applications. The proposed architecture integrates multi-parameter sensing, embedded edge processing, wireless communication, cloud-based data management, and real-time visualization within an Internet of Things (IoT) framework. Key environmental and electrical parameters, including solar irradiance, temperature, voltage, current, and power output, are acquired, processed, and transmitted for continuous system evaluation. The performance of the RTSDAS is assessed using simulated operational scenarios to examine signal quality, measurement accuracy, communication effectiveness, system efficiency, and fault detection capability. Results demonstrate that the implemented filtering techniques effectively reduce measurement noise while preserving signal integrity, leading to more reliable data acquisition. The PV power output accurately follows irradiance variations, with the system achieving a peak operational efficiency of approximately 98% under optimal conditions. Furthermore, the fault detection mechanism successfully identifies abnormal operating conditions in real time, enabling prompt maintenance intervention and improved system reliability. The findings confirm that the proposed RTSDAS provides a robust, scalable, and cost-effective solution for solar energy monitoring and is well suited for deployment in smart grid, microgrid, and distributed renewable energy environments.

IoT-based Water Quality Analysis and Fish Species Detection using Deep Learning

2026-05-11T08:21:30+00:00

Water quality plays a vital role in maintaining aquatic ecosystems and ensuring sustainable aquaculture operations. Continuous monitoring of key parameters such as pH, temperature, and dissolved oxygen is crucial to maintain the optimal environment for aquatic life and to prevent health hazards in fish farming systems. Traditional water quality monitoring methods often involve manual sampling and laboratory analysis, which are time-consuming, less accurate, and fail to provide real-time insights into the aquatic environment. This project presents a smart IoT-based water quality monitoring and fish species detection system integrated with deep learning technology. IoT sensors continuously collect and transmit data related to pH, temperature, and dissolved oxygen to a cloud server for real-time visualization and analysis. Simultaneously, a deep learning model identifies and classifies fish species using image data, enabling intelligent management of aquaculture resources. The system enhances productivity, minimizes human error, and supports sustainable aquaculture by ensuring timely intervention through automated alerts and predictive insights.