The work of (Mustafa & Al-Saif, 2014) and the

The model is simulated and validate with different range of joint angles as indicated in fig. (4) and (5). In figure (4), the initial joint angles for arm 1 and arm 2 are  and  respectively and the desired joint angle positions we want the arm to reach are  and . The control strategy ensures that the desired joint angle positions are obtainable by selecting suitable controller gains as indicated in table (2). It can also be deduced that arm1 and arm 2 followed the desired trajectory as indicated with red and blue line respectively. Furthermore, different angle conditions are selected to ensure that the robot arm performs efficiently, so the initial joint angle of arm 1 is  and that of arm 2 is  and we expected that the desired joint angle positions should be at  and  as indicated in figure (5). It can be deduced that for the robot arm to reach the desired trajectory (angle position), the gains of the PID controller need to be adjusted at every instant and tuned to prevent overshoot and oscillation that associated with changing of parameter values. It can also be noticed that the torque applied as shown in the figure (6) and figure (7) slightly overshoot but stabilize quickly. It can be observed that the parameters values influence the controller performance, so adequate online auto tuning of the parameters of the controller will enhance the parameters selection. The model is validated with the work of (Mustafa & Al-Saif, 2014) and the result obtained from the nonlinear model show a similar responses with the  results presented in this paper. However, the PID decoupling approach adopted in the work of (Mustafa & Al-Saif, 2014) is too rigorous and limited to specified range of joint angles but the method presented in this research work permit flexibility of joint angles selection and decoupling is dependent on the Jacobian matrix derivation.