Design and Implementation of Fully Automated Calibration System for Multiple Temperature-Controlled Enclosures Using Networked Measurement and Real-time Steady-State Detection Approaches

摘要

This paper presents the system design, algorithms and user applications for the temperature-controlled enclosure calibration and testing processes. The system works as an IoT-based real-time temperature steady-state detection and monitoring system. The main goals of this research are divided into four parts. The first one is to design a networked calibration system that can perform multiple devices calibrating and testing at the same time. The system utilizes Modbus and Internet-based Wireless communication networks. The second one is to study and implement an appropriate algorithm that can detect the temperature steady-state of the temperature-controlled enclosure devices connected in the network. This proposed algorithm, the steady-state detector is implemented based on a low-pass filter and finite-state machine approaches. It is implemented using TypeScript, the strongly typed JavaScript. The third one is to design and implement web and mobile that supported applications used as real-time data visualization and user interface. The application is employed with development languages, including HTML, CSS, and TypeScript. The last one is to design and develop a web-based application to evaluate and deliver a technical report of the calibration and testing. The TypeScript files are exploited for the steady-state detector and applications are transpiled to common JavaScript that can perform directly in all browsers. The proposed system is evaluated in the real-use environment with various possible conditions. The experimental results validate the effectiveness of both system architecture, algorithm and user applications, offering overall performance comparable with legacy manual calibrating, testing and reporting processes. In particular, the normalized error (En) between manual and the proposed system significantly different (En> 1), namely 95% confidence interval, thus widely satisfying requirements imposed by standardization bodies.