PREDICTION OF SUPPORTING HAND FORCES FOR COMMON AUTOMOTIVE ASSEMBLY TASKS BASED ON OPTIMIZATION AND STABILITY TECHNIQUES FOR GIVEN POSTURE

摘要

Quite often people are faced with one handed tasks in which the other hand is needed for support. Without these supporting external forces, postures may be unstable, rendering the task impossible. Automotive assembly line operators are confronted with these types of tasks every day, such as hose installations and the connection of electrical components. Determining the optimal location and forces for the supporting hand is important to minimize potential injuries of operators. Traditionally, these supporting hand forces are measured by experiments. This work attempts to provide an important predictive tool that promises to be of considerable value to companies in predicting leaning forces in work simulation for the proactive ergonomic assessment of work tasks. It presents a method using optimization and stability analysis techniques. Stability is based on the calculation of a three dimensional zero moment point (3D-ZMP) and the resultant reaction loads, calculated from the joint torque. The formulation of the optimization problem used to predict the supporting hand forces is presented and tested using tasks commonly encountered by automotive assembly workers. The results are compared to that in literature, providing an initial validation of the methods. The predicted external forces fell within the 95% confidence intervals calculated from the literature for all tasks.