Diagnosing Equipment Malfunctions Using Light Indication and Sensors

Author Name(s): Rustem R. Ziyatdinov, Dmitriy A. Lobachev, Aleksandr A. Shabaev, Ramil T. Nasibullin
Author Email: rust.kfu@gmail.com

Abstract

Diagnostics of production equipment is an urgent task for modern enterprises. It allows timely identifying the emerging equipment malfunctions, prevent their occurrence and switch from emergency repairs to scheduled preventive repairs. One of the ways to identify a malfunction is to visually inspect the equipment itself, use the instrumentation readings and the light indication of equipment control systems. Moreover, the presence of a large amount of information causes difficulties when the maintenance personnel conducts diagnostics and requires high qualifications. To solve this problem, one can use the image recognition and the formation of recommendations for maintenance and repair using an intelligent system. Image recognition is one of the most sought-after functions of modern control systems used in various activity fields. This article proposes a malfunction diagnosing method by light indication, based on the use of artificial intelligence elements. The input information is represented by the images of the status of light indicators of the equipment and the instrumentation readings. The resulting image is pre-processed and classified in order to determine the equipment condition. The correlation of the light indication image to a certain state of equipment is carried out using one of the classification algorithms. The most promising classification methods in image recognition problems can be considered the methods of support vectors and k – the nearest neighbors. Further, on the basis of the received equipment condition, recommendations for its maintenance and repair are formed with the help of an expert system. To develop an expert system, we used a production model that allows applying simple and accurate mechanisms for using knowledge. Thus, an intelligent system will allow diagnosing production equipment faster and more efficiently, without imposing high requirements on the staff qualifications.

Introduction

Increased requirements for the quality of products require continuous improvement of production technology. One way to solve this problem is the integrated enterprise automation. At the same time, it is important to carry out timely equipment diagnostics for the transition from emergency repairs to scheduled preventive repairs [1, 2]. At the same time, the variety of used production equipment and automation means creates additional difficulties in integrating all the existing subsystems into a single system. One of the possible directions for solving such problems is the use of intelligent systems with the elements of technical vision. Previously, the development of such systems was constrained by high demands on computing resources. The emergence of new image processing and classification techniques, increased productivity of modern computing systems removes almost all the problems of system implementation based on technical vision. Image recognition problems are encountered in a wide variety of fields: from fault detection in diagnosing systems to driving unmanned vehicles [3]. Management of production equipment is carried out using industrial controllers. Information at the inputs and outputs of industrial controllers determines equipment condition. Therefore, to diagnose the condition and malfunctions of industrial equipment, one can use the light indication on the input/output modules of industrial controllers. At the same time, the light indication on the industrial controller can have hundreds of indicators, and when maintenance personnel detects and diagnoses a malfunction, difficulties arise in the data interpretation. In addition, some malfunctions may require in-depth theoretical knowledge and extensive practical experience. Expert systems can be used to solve the problem of diagnosing equipment malfunctions [4, 5]. The task of the expert system in this case is to identify equipment malfunctions according to the input data and issue of recommendations for maintenance or repair, if necessary.

Conclusion

The system allows determining the status of production equipment and issue recommendations for its repair based on the input and output signals of its control system. At the same time, the speed of the equipment diagnosing process increases many times over in comparison with the time spent on diagnosing malfunctions by maintenance personnel. This system can be easily scaled and adapted to a wide range of different equipment. Another feature of the system use is that the diagnosing process itself can be performed by low-skilled service personnel. All this allows reducing equipment downtime due to malfunctions and increase production efficiency

599 total views, no views today

Download PDF File

About the author: admin