2. Increasing the weighting ratio value
J=
T has a
beneficial effect in reducing the jerks but lengthens
the execution time. Therefore, the total execution
time is necessary in the cost function.
3. The vector of torque is derived as the inertia matrix
multiplied by the vector of jerk (i.e. s_ ¼ I J). It is
easy to infer that maximizing the value
J=
T helps to
reduce the jerks which in turn reduces the jumps of
torque.
Figure 16 shows the convergence of cost function
versus iterations. A continuous decrease in average
of cost function is also indicative of a smooth
convergence to a solution. As seen in Figure 16,
when
J=
T ¼ 0.05, the time-jerk synthetic optimal
trajectory is the best one comparing with other
cases.
Figure 17 shows the total control energy of the
manipulator for different kinds of weighting ratios.
It is obvious that the control input requires relative
minimum energy comparing with other cases