40 -


 20


 i -
 0 0
 0 0


 -20


 -40
 0 60 120 180 240 300 360 420
 Time (1/60 sec)

Figure 9.5. Typical closed-loop response of joint 1 to an alternating step input of 32.6 deg/sec,
 joint 2 extended.


input steps of 48.8 deg/sec (Figure 9.6), the system responded with average steady-state

velocities of 43.6 deg/sec (11 percent) and -44.3 deg/sec (9 percent). For this case, joint 2 was

centered in its support, and no overshoot was observed. In summary, the lag-lead velocity

controller as implemented for joint 1 achieved very good results. With large input step values,

steady-state errors of no more than 12 percent and overshoot values of no more than 18 percent

were observed. These values fell well below the maximum allowable errors as specified as

performance requirements (40 percent allowable steady-state error and 45 percent allowable

overshoot).

 Unlike the uses of the velocity controllers for joints 0 and 1, the velocity controller for

joint 2 was used during fruit removal. Velocity control of joint 2 was used for extending the end-

effector toward a targeted fruit. Also, the velocity error of joint 2 was used for detection of

collisions and unremovable fruit. Therefore the performance of the joint 2 velocity controller

was much more critical than that of the other two joints. Typical step response plots for joint 2

velocity controller are shown in Figures 9.7 and 9.8. The response of the system to two

individual step inputs is presented in Figure 9.7. In the first step, a velocity setpoint of