Online Identification for the Automated Threaded Fastening Using GUI Format

Abstract

Screw insertions are a common joining process and are especially popular in assemblies that need to be disassembled for repair, maintenance or relocation. A human performing screw fastening will typically use four stages (L.D. Seneviratne et al., 1992): 1) Alignment: the holes in the mated parts are aligned and firmly held together. 2) Positioning: the screw is globally positioned with respect to the aligned holes. 3) Orientation: the screw is oriented until its axis coincides with the axis of the aligned holes. 4) Turning: the screw is turned with the appropriate torque until fastening is achieved. In manual screw insertions, the torque exerted by the screwdriver depends mainly on the applied operator force. Human operators are particularly good at on-line monitoring of the operation. However, with power tools, the increased insertion speed reduces the human ability to monitor the insertions on-line. Thus on-line automated monitoring strategies for the screw fastening process are highly desirable. One such approach is based on the “Torque Vs. Insertion Depth” signal measured in real time; if this signal is within a pre-defined bound of the correct insertion signal, then the insertion is considered to be satisfactory. The torque signature signal for a correct insertion is either taught as predicted using an analytical model (Klingajay & Seneviratne, 2002). Industrial applications of automated screw insertions have been implemented in several forms. These achieved forms are applied in different objectives. With the development of electrically powered screwdrivers the attempts at automating the screw insertion process with emphasis on the torque signature signal vs. angle signal and become to the primary mathematical model. In 1997, this analytical model was implemented by (Ngemoh, 1997). The Neural Network techniques have been applied by using the ability of Weightless to monitor the screw insertion processes in difference insertion cases (Visuwan, 1999). Bruno has distinguished between successful and unsuccessful insertion based on Radial Basic function (Bruno, 2000). Both monitoring performs are to apply Artificial Neural Network in view points of classifications. “A distinction without a difference has been introduced by certain writers who distinguish ‘Point estimation’, meaning some process of arriving at an estimate without regard to its precision, from ‘Interval estimation’ in which the precision of the estimate is to some extent taken into account” (Fisher, 1959). Fisher founded the Probability theory as logic agree, which gives us automatically both point and interval estimates from a single calculation. The distinction commonly made between hypothesis