Integrated Wireless Technologies with Computer for Industrial Machinery Fault Diagnosis: Challenges Comparison and Characteristics: A Review

Moneer Ali Lilo
2018 Muthanna Journal of Pure Science  
Condition monitoring of machinery in industries is becoming an emerging technique for efficient operations and productivity. A wide range of fault diagnosis approaches have been proposed to improve machinery operations in industries. These monitoring and fault diagnosis approaches are most effective for reliable machinery operations. However, due to the harsh environment of industries, these approaches have been suffered from different challenges. In order to extract the significant features
more » ... fault diagnosis and monitoring, different neural, fuzzy, and signal processing based systems were adopted. In this paper, we discuss wireless sensor based fault diagnosis and monitoring approaches and their types for industrial machinery. Furthermore, the paper presents industrial challenges to adopting these approaches and related efforts. Several different types of WSN based solutions have been proposed to address these constraints. Most of WSN applications have been used for data acquisition and transmission and extract the fault diagnosis functions. These types of diagnosis are promising alternative approach to transmitting the raw data and reduce the quality and save energy and time. In many industrial WSN applications, it is difficult to obtain all information about the device and their faults by a single sensor node. Due to the harsh environment of industries, the wireless-based devices have been suffered in communication quality in terms of noise and interference. These issues disturb the communication quality and increase uncertainty in diagnosis results. Data fusion was presented to augment the accuracy of transmitting process based on combining different types of data into one packet. It is effective method to save energy and computation power such as fuzzy data fusion [1], Baysian method [2], Dempster-Shafer theory [3] . These methods are based on wired sensor systems and have some limitations.
doi:10.18081/2226-3284/018-6/61-75 fatcat:7bc435gq4bfrjep5vyd7egfbhy