Journal of Intelligent & Fuzzy Systems 29 (2015) 2447–2455 2447

DOI:10.3233/IFS-151945
IOS Press

Fuzzy PID control method of deburring

industrial robots
Yong Taoa,∗ , Jiaqi Zhengb , Yuanchang Linc , Tianmiao Wanga , Hegen Xiongb , Guotian Hec
and Dong Xua
a BeihangUniversity, Beijing, China
b Wuhan University of Science and Technology, Wuhan, China
c Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences, Chongqing, China

Abstract. This paper proposes a fuzzy PID control method for deburring industrial robots. The adaptive fuzzy PID controller
relates to the trajectory and joint angular parameters of the end-effector on a robot. The PID controller parameters update online
at each sampling time to guarantee trajectory accuracy of the end-effector. The simulation of the fuzzy PID control is provided
based on the 6-DOF deburring industrial robot. Experimental results demonstrate the efficiency of the fuzzy PID control method.

Keywords: Fuzzy PID controller, robot trajectory, joint angular, deburring industrial robot

1. Introduction material should also be taken into account. An adaptive

modeling method based on statistic machine learning
Industrial robots are used in industrial applications was proposed [18], as was an intelligent control method
such as welding, grinding, and painting. It is therefore for the industrial robot was proposed [19]. A nonlinear
important that end-effector of such robots moves accu- damping control scheme that reduces the force over-
rately and smoothly to achieve the desired trajectory. shoot without compromising the dynamic response was
This paper studies the aluminum casting of motorcy- proposed [2]. A solution of unmanned helicopters using
cle accessories. Aluminum burrs are of irregular shapes a novel nonlinear model predictive controller enhanced
and large volumes. Workers typically remove burrs, but by a recurrent neural network has also been proposed
this process can lead to injuries if the artificial manu- [7].
facturing method is out of date. A new methodology based on a multi-objective
This paper proposes an industrial deburring robot, genetic algorithm was proposed [12]. An algorithm
which is safer and more efficient than the manual designed for high-accuracy local dynamic identification
removal of burrs. The burr removal process is com- was presented [10]. A novel method to study precision
plicated because each burr is unique, and therefore grip control using a grounded robotic gripper with two
requires special steps. Some experts suggest the use of moving, mechanically-coupled finger pads equipped
the force/position control strategy to guarantee product with force sensors was presented [11]. A fuzzy rein-
accuracy [6]. This paper introduces an effective online forcement learning (FRL) scheme based on fuzzy logic
force control direction adjustment algorithm using the and the sliding-mode control was presented [13].
fuzzy vector method. The position and orientation of the end-effector are
The workpiece and tool should be taken into account important for the deburring robot. The end-effector ori-
during manufacturing. An adaptive fuzzy hybrid posi- entation and position freedom, as well as the robot
tion/force controller was proposed [4]. The removed configuration, greatly influences the cost of support-
∗ Corresponding author. Yong Tao, Beihang University, 37
ing movement [14]. The design and implementation of
Xueyuan Road, Beijing, China. Tel.:/Fax: +86 10 82328271; E-mail: a platform for rapid external sensor integration into an
taoy@buaa.edu.cn. industrial robot control system was presented [1].
ISSN 1064-1246/15/$35.00 © 2015 – IOS Press and the authors. All rights reserved
This article is published online with Open Access and distributed under the terms of the Creative Commons Attribution Non-Commercial License.
2448 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots

The intelligent control methodology is used to solve

sophisticated tasks. The intelligent controller work
condition, which takes into account human expert
knowledge and skills for planning and controlling
sophisticated tasks, was introduced [8, 16].
A method of path planning was introduced for auto-
matic grinding robotic blade. The robot program, which
contains the grinding path generated by a PC, was
demonstrated [20]. A controller based on a GA-fuzzy-
immune PID, a robot dexterous hand control system,
and a mathematical model of an index finger were also
presented [17]. Co-simulations of a novel exoskeleton-
human robot system based on humanoid gaits with
Fuzzy PID algorithms was presented [3]. Fig. 1. The deburring industrial robot system.
Vibration suppression in a two-inertia system using
an integration of a fractional-order disturbance observer Table 1
and a single neuron-based PI fuzzy controller was D-H parameters of the manipulator
presented [15]. A collocated proportional-derivative Link i Joint angle αi−1 / ai−1 /mm di /mm Range of
(PD)-typeFuzzyLogicController(FLC)wasdeveloped, θi /(◦) (◦) joint angle
which was used to study controller effectiveness [9]. The 1 θ1 0 0 d1 ±180◦
hierarchical fuzzy logic is provided according to the sin- 2 θ2 –90 a1 0 +80/–135◦
gular perturbation approach, which can derive the slow 3 θ3 0 a2 0 +225/–80
4 θ4 –90 a3 d4 ±360◦
and fast subsystem of a flexible-link robot arm control 5 θ5 90 0 0 ±120◦
[5]. 6 θ6 –90 0 0 ±450◦
Traditional PID parameters could not be modified.
However, the advantages of the PID control method are 2.1. Manipulator kinematics
the easy measurement of parameters, which is helpful
in choosing optimal values. Manipulator kinematics involve geometric and time-
The fuzzy rules of the fuzzy control strategy are dif- based motion properties, particularly the interactions
ficult to determine. The parameters of the Fuzzy PID between the various links over time. In this paper,
strategy can be updated online according to the practical the manipulator kinematics are described by the D-H
conditions. In this paper, PID parameters are modified method. The six links are built in separate coordi-
according to actual working conditions. Robot control nates; their positions and orientations are described in
efficiency and accuracy is improved based on the fuzzy the coordinates. Table 1 describes the four parameters
PID algorithm. The paper is organized as follows: Sec- ai−1 , αi−1 , di , θi .
tion 2 introduces the robotic system and manipulator
kinematics; Section 3 presents the traditional PID con- 2.2. Kinematics model based on D-H notation and
troller and mathematical model; Section 4 details the ER50-C20 link parameters
fuzzy control approach; Section 5 describes the fuzzy
PID controller design and improvements; Section 6 There are some relative parameters on the ER50-
compares the simulations between the traditional and C20:
fuzzy PID controller; and Section 7 summarizes the a1 = 220 mm, a2 = 900 mm, a3 = 160 mm, d1 =
conclusions. 563 mm, and d4 = 1014 mm.
According to the inverse kinematics solution method,
2. The introduction and model the angular joint of the end-effector can be calculated
of the robot system as follows:
Figure 1 shows the 6-DOF deburring industrial robot. θ6 = A tan 2(ox cθ1 s(θ2 + θ3 ) + oy sθ1 s(θ2 + θ3 )
It is composed of a base, waist, arm, and manipulator.
+oz c(θ2 + θ3 )) − nx cθ1 s(θ2 + θ3 )
The robot kinematic model is described by homoge-
neous coordinates. −ny sθ1 s(θ2 + θ3 ) − nz c(θ2 + θ3 )) (1)
 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots 2449

θ2 = A tan 2(s(θ2 + θ3 ), c(θ2 + θ3 )) − A tan 2(a3 , d4 )

+A tan 2(k, ± a32 + d42 − k2 )


θ3 = A tan 2(a3 , d4 ) − A tan 2(k, ± a32 + d42 − k2 )

where cθi represents cos θi , cαi represents cos αi , sθi Fig. 2. The traditional PID based on position feedback.
represents sin θi , and sαi represents sin αi . Additionally,
Therefore, the transfer function is obtained as
k = (m2 + n2 − 9002 − 1602 − 10142 )/(2 × 900) follows:
m = cθ1 px + sθ1 py − 220 u(t) Ki
G(s) = = Kd s + Kp + (4)
n = pz − 563 e s
As shown in Fig. 2, a desired trajectory is obtained
The manipulator kinematics were discussed and the by the teaching method, and the desired position of the
kinematic model based on D-H notation was built. The end-effector Xd is obtained. The desired joint angular θd
relationship between the position p = (px , py , pz ) of of the end-effector is calculated by inverse kinematics,
the end-effector and the angular joint θ is obtained. as described by Equation (1). The desired joint angular
According to the known end of the link relative to θd is output by the PID controller and the robot model.
the position and orientation of the reference coordi- This represents the teaching method process. When the
nate system, the inverse kinematics are used to calculate second workpiece is deburred, the value e = (θd − θ)
each joint variable. The error between the desired angu- is regarded as the input. The error e is the output of
lar joint and actual angular joint is regarded as input. the PID controller and robot model. The optimal actual
The following sections describe the PID mathematical joint angular θ is calculated, and the actual joint angular
model. θ is recorded by a coded disk. The next workpiece is
manufactured in this way.
Though the traditional PID controller could detect the
3. The PID controller model in a deburring parameter values of the end-effector, the shortage of the
robot traditional PID algorithm is evident because its param-
eters cannot self-adjust to control complex, nonlinear
In an ER50-C20 robot, the machining path of the subjects; thus, this method may create errors when the
following workpieces is machined according to the current workpiece is different from the previous one.
machining path predetermined by the related numerical Fuzzy PID can be applied to this robot, enabling it to
value of the terminal trajectory, recorded by a coded cope with the difficulty of complicated nonlinear sys-
disk and the artificial teaching method. The optimal tems. A fuzzy PID algorithm can self-adjust parameters
machining path can be achieved by proper adjustments according to rules designed by experience.
to meet the accuracy requirements. If the desired tra-
jectory is known, the PID controller is placed into the
above linear system. 4. The fuzzy PID control approach

u(t) = Kd (θ̇d − θ̇) + Kp (θd − θ) + Ki The fuzzy controller can measure the outputs of the
  traditional PID controller and alter them in order to
(θd − θ)dt = Kd ė + Kp e + Ki edt (2) reduce output errors. The fuzzy controller can make
accurate decisions to improve machining accuracy
where based on the knowledge of human operation and expe-
rience. The fuzzy controller has integral fuzzy sets,
e = θd − θ; (3) which are designed by work conditions; these fuzzy
sets are small, medium, large, etc. The fuzzy sets can
e is the joint angular error; ė or ec is the joint angular be converted into the automatic control strategy based
error rate; θd is the desired joint angular; and θ is the on a knowledge base. The fuzzy controller consists
actual joint angular. of fuzzification, fuzzy reasoning, defuzzification, and
2450 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots

the knowledge base. When the fuzzy sets have been signals and feedback signals are controlled by the fuzzy
designed, the inputs must adapt to the correspond- PID controller. Figure 8 shows a detailed fuzzy PID con-
ing rules. The output values can be obtained by fuzzy troller structure; the fuzzy rules are shown as Equations
reasoning based on fuzzy rules; better output values (5). The fuzzy subset is {NL, NM, NS, ZE, PS, PM,
acquire better results. Finally, inputs have m fuzzy val- PL} and the fuzzy domain is {–0.3, –0.2, –0.1, 0, 0.1,
ues; the number of rules are represented by mn , where 0.2, 0.3}, in which the NL, NM, NS, ZE, PS, PM, and
n is the number of inputs. PL stand for negative large, negative middle, negative
small, zero, positive small, positive middle and positive
large, respectively. The value of e represents the error
5. The fuzzy PID control algorithm and ec represents the error rate. The triangle-type mem-
bership function is adopted for each fuzzy variable, as
5.1. The design of the fuzzy PID controller shown in Figs. 4–6, which show the membership func-
tion curves of e, ec, dKp , dKi , dKd , respectively.
The fuzzy PID controller meets e and ec requirements The number of fuzzy sets is chosen based on the
for PID parameters at different times. Figure 3 depicts desired manufacturing needs and workpiece quality.
the basic process. In general, there is no specific method for designing
The fuzzy PID control system is composed of the
input, output, and feedback loop. The error and error
rate of the joint angular of the end-effector are regarded
as the inputs, and the three parameters Kp , Ki , Kd
represent the outputs. As shown in Fig. 7, the con-
troller consists of the traditional PID controller and
fuzzy controller. The three parameters Kp , Ki , Kd of
the traditional PID controller are updated online by the
fuzzy rules. The fuzzy rules are designed in the fuzzy
controller. The values e = θd − θ and ec = θd − θ  are
regarded as the inputs; these are exported by the con-
troller and robot model. The joint angular error e Fig. 4. Membership function curves of e and ec.
and ec are placed into the fuzzy controller; the com-
pensation amounts Kp , Ki , Kd are exported by
the fuzzy controller. The values of Kp , Ki , Kd
are regarded as the inputs of the traditional PID con-
troller to adjust the values of Kp , Ki , Kd . The
optimal three parameters Kp , Ki , Kd of the tra-
ditional PID controller are adjusted online by the fuzzy
controller.
The proper actual joint angular θ is obtained. In each
control loop, the error and error rates between the input
Fig. 5. Membership function curves of dKp and dKi .

Fig. 3. The fuzzy PID algorithm flow chart. Fig. 6. Membership function curves of dKd .
 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots 2451

properly applied to the traditional PID controller and the

fuzzy PID controller. The fuzzy rules should be clearly
classified in order to increase effectiveness. However,
too many rules will have a negative impact on sys-
tem stability. According to the work condition, if the
fuzzy sets have less than two rules, the fuzzy subset
will become {NL, NS, ZE, PS, PL}. Because the work-
piece is shaped irregularly, two rules are not able to
Fig. 7. The fuzzy PID controller. meet the change requirement, which should be vari-
able based on the blur amount. However, if there are
too many rules, the end-effector stability will decrease.
Therefore, three fuzzy sets are more reasonable. The
fuzzy subset is {NL, NM, NS, ZE, PS, PM, PL}.

5.2.1. Design of control rules of Kp

In the PID system, the response speed and steady-
Fig. 8. Working principle of the fuzzy controller. state error are affected by the value of Kp . At the
beginning of the regulation, the value of Kp is increased
to obtain faster response speeds. In the middle of the
the fuzzy controller, but input and output membership regulation, the smaller overshoot and faster response
functions to emulate human decision-making can be speed must be considered, and the value of Kp
designed. The following methods are typically used to decreases. Increasing Kp at the end of the regulation
design a fuzzy controller: promotes machining precision. Kp is the compen-
– Model the knowledge of the control engineer. sation amount. The fuzzy domain of Kp is {–0.03,
– Model of the actions and experience of the human –0.02, –0.01, 0, 0.01, 0.02, 0.03}, which corresponds
operator. to {NL, NM, NS, ZE, PS, PM, PL}.
– Fuzzy model of the controlled plant.
As shown in Fig. 8, the error between the desired 5.2.2. Design of control rules of Ki
joint angular θd and the actual joint angular θ of the In the PID system, the integral control eliminates the
end-effector is regarded as the input, noted by e. The steady-state error of the system. At the beginning of the
value of e is fuzzy according to the fuzzy controller, regulation, Ki is reduced to prevent integral saturation.
used in fuzzy operators. The process is fuzzy reason- In the middle of the regulation, Ki is increased to guar-
ing, which obtains the optimal fuzzy value. The optimal antee system stability. At the end of the regulation, the
fuzzy value should be defuzzified to obtain the clear static error is reduced by increasing Ki . Ki is the com-
value, which is used in the mathematical model of the pensation amount. The fuzzy domain of Ki is {–0.03,
robot. –0.02, –0.01, 0, 0.01, 0.02, 0.03}, which corresponds
to {NL, NM, NS, ZE, PS, PM, PL}.
5.2. Design of control rules of Kp , Ki , Kd

Three positive trapezoidal fuzzy sets are required to 5.2.3. Design of control rules of Kd
cover the input and output variables. In the PID system, the differential control can change
The material of the workpiece is aluminum and the the dynamic characteristics. At the beginning of the
material of the tool is cast iron. The analysis of force regulation, Kd is increased to prevent overshooting. In
between the tool and the workpiece should be consid- the middle of the regulation, Kd is reduced to a constant.
ered. Combining the experience of the workers and At the end of the regulation, Kd is further decreased
work conditions, three different fuzzy rules for each to reduce the settling time. Kd is the compensation
gain are utilized, and the three parameters Kp , Ki , Kd amount. The fuzzy domain of Kd is {–0.3, –0.2, –0.1,
can be designed based on the analysis. Test results 0, 0.1, 0.2, 0.3}, which corresponds to {NL, NM, NS,
indicate that Kp = 4, Ki = 0.001, Kd = 100 could be ZE, PS, PM, PL}.
2452 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots

5.3. The rules of fuzzy control Table 4

Fuzzy rule table for change in Kd
(1) When |e| is a greater value, Kp must also be e/ec NL NM NS ZE PS PM PL
bigger, Kd should be smaller, and Ki = 0. NL 0.1 –0.1 –0.3 –0.3 –0.3 –0.2 0.1
(2) When |e| is a middle value, Kp should NM 0.1 –0.1 –0.3 –0.2 –0.2 –0.1 0
NS 0 –0.1 –0.2 –0.1 –0.1 –0.1 0
decrease, and Kd and Ki should be of middle ZE 0 –0.1 –0.1 –0.1 –0.1 –0.1 0
value. PS 0 0 0 0 0 0 0
(3) When |e| is a smaller value, Kp and Ki PM 0 0.1 0.1 0.1 0.1 0.1 0.3
PL 0.3 0.2 0.2 0.2 0.1 0.1 0.3
should be greater, and Kd should be a middle
value.
The rules of the Fuzzy PID controller are as follows: to the fuzzy rule tables, the PID parameters are modified
as follows:
If e is NL and ec is NL, then Kp is PL, Kp = Kp + Kp
Ki is NL, Kd is PS.
Ki = Ki + Ki (6)
If e is NM and ec is NL, then Kp is PL,
(5) Kd = Kd + Kd
Ki is NL, Kd is PS.
The output of the fuzzy PID algorithm are as follows:
If e is PL and ec is PL, then Kp is NL, 
  
Ki is PL, Kd is PL. u(t) = Kd (θ̇d − θ̇) + Kp (θd − θ) + Ki (θd − θ)dt

There are 49 rules in the fuzzy logical controller, = (Kd + Kd )ė + (Kp + Kp )e
as shown in Tables 2–4. In this paper, 49 fuzzy rules 
were considered necessary and created to implement +(Ki + Ki ) edt (7)
the proposed fuzzy PID controller for the deburring
robot system. The PID parameters were updated online Kp , Ki , Kd are improved values that are modified by
according to these rules. The inputs of the fuzzy logic the fuzzy PID controller. Kp , Ki , Kd are the original
block are the angular displacement error e and deriva- values.
tive of error ec. The inputs are modified by the fuzzy PID
feedback system to optimize the trajectory. According 5.4. Output defuzzification

Table 2 The center average method was used for defuzzifica-

Fuzzy rule table for change in Kp tion. The output u is calculated as:
e/ec NL NM NS ZE PS PM PL 
n
NL 0.03 0.03 0.03 0.03 0.02 0.01 0 dui µAi
i=1
NM 0.03 0.03 0.03 0.03 0.02 0 0 du = (8)
NS 0.02 0.02 0.02 0.02 0 0.01 0.01 n
ZE 0.02 0.02 0.01 0 –0.01 –0.01 –0.01
µAi
i=1
PS 0.01 0.01 0 –0.02 –0.02 –0.02 –0.02
PM 0.01 0 –0.01 –0.02 –0.02 –0.02 –0.03 where du = {dKp , dKi , dKd } represents the centers
PL 0 0 –0.02 –0.02 –0.02 –0.03 –0.03
of the output membership functions; µA1 represents the
centers of the input membership function e; µA2 repre-
Table 3 sents the centers of the input membership function ec;
Fuzzy rule table for change in Ki and n represents the input number.
e/ec NL NM NS ZE PS PM PL If the input is known, the output can be calculated
NL –0.03 –0.03 –0.02 –0.02 –0.01 0 0 by Equations (8). Combining Equations (7) and (8),
NM –0.03 –0.03 –0.02 –0.01 –0.01 0 0 Kp , Ki , Kd are obtained to control the end-effector
NS –0.03 –0.02 –0.01 –0.01 0 0.01 0.01
ZE –0.02 –0.02 –0.01 0 0.01 0.02 0.02 trajectory. Kp , Ki , Kd are adjusted by the fuzzy PID
PS –0.02 –0.01 0 0.01 0.01 0.02 0.03 under actual processing conditions. The simulations in
PM 0 0 0.01 –0.02 0.02 0.03 0.03 the next session verify the feasibility of the traditional
PL 0 0 0.01 0.02 0.02 0.03 0.03
PID and the fuzzy PID algorithm.
 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots 2453

6. Experiments and simulation analysis

The methods described in this paper have been tested

in simulations of the ER50-C20 robot. Given the orig-
inal value of joint θ = [0.176, –0.701, 0.079, –0.0002,
–0.949, –1.019] and the final value of joint θ  = [–0.236
–0.166 –0.691 0 –0.715 –0.237]. The end-effector of the
robot moves from point A (1350, 530, 1250) to point
B (1615, 240, 490) in Cartesian space. A total of 101
points were sampled between A and B. The sample time
was 5 seconds, sampled once every 0.05 seconds. Fig. 11. Error of joint angular based on traditional and fuzzy PID.
Figure 9 shows the trajectory curves simulated by
the Matlab toolbox; the curves are formed by interpo-
lation. The end-effector trajectory is determined by the
artificial teaching method, as shown in Fig. 9, which
also indicates the three dimensional trajectory of the
end-effector.
The traditional PID method requires a long setting
time of approximately 1s, as shown in Fig. 10. The
error curve is shown in Fig. 11. The red solid line rep-
resents the fuzzy PID control method; the black dotted
line represents the traditional PID control method. The
fuzzy PID algorithm requires a short setting time of Fig. 12. The value of Kp based on Fuzzy PID.
0.3s, as shown in Fig. 10. Figure 11 depicts the corre-
sponding error curve. The fuzzy PID control strategy
error value is approximately 10–4 rad in Fig. 11. The
error value of the traditional PID control method is

Fig. 13. The value of Ki based on Fuzzy PID.

Fig. 9. The orientation of the end-effector. approximately 0.01rad, as shown in Fig. 11. This indi-
cates that the fuzzy PID controller is more accurate than
the traditional PID controller. Comparing the red solid
line to the black dotted line in Fig. 10, the setting time
of the black dotted line is more than 3 times that of
the red solid line. Thus, the fuzzy PID controller has a
shorter setting time than the traditional PID controller.
The fuzzy PID overshot by approximately 7.5%, which
is within the permitted 20% scope. The overshot value
has small impact on the workpiece, and results meet the
requirements.
According to the simulation, the setting time is
greatly reduced and the joint angular error is decreased
Fig. 10. Joint angular based on traditional and fuzzy PID. for the fuzzy PID controller. The fuzzy PID controller
2454 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots

ring, IEEE International Conference on Intelligent Robots and

Systems, 2006, pp. 2743–2748.
[2] C.Y. Lai, Improving the transient performance in robotics
force control using nonlinear damping, IEEE International
Conference on Advanced Intelligent Mechatronics, 2014, pp.
892–897.
[3] D. Pan, F. Gao, Y.J. Miao and R. Cao, Co-simulation research
of a novel exoskeleton-human robot system on humanoid gaits
with Fuzzy PID/PID algorithms, Advances in Engineering
Software 79 (2015), 36–46.
[4] F.Y. Hsu and L.C. Fu, Intelligent robot deburring using adap-
tive fuzzy hybrid position/force control, IEEE Transactions on
Robotics and Automation 16(4) (2000), 325–335.
Fig. 14. The value of Kd based on Fuzzy PID. [5] J. Lin and F.L. Lewis, Two-time scale fuzzy logic controller of
flexible link robot arm, Fuzzy Sets and Systems 139(2) (2003),
performs better than the traditional PID algorithm. As 125–149.
shown in Fig. 12, Kp is updated online by the fuzzy [6] K. Kiguchi and T. Fukuda, Position/force control of robot
manipulators for geometrically unknown objects using fuzzy
PID algorithm, and the optimal value of Kp is obtained neural networks, IEEE Transactions on Industrial Electronics
according to the actual processing conditions. As shown 47(3) (2000), 641–649.
in Figs. 13 and 14, the values of Ki and Kd are updated [7] K. Dalamagkidis, K.P. Valavanis and L.A. Piegl, Nonlinear
online. The fuzzy PID controller values of Kp , Ki , Kd model predictive control with neural network optimization for
autonomous autorotation of small unmanned helicopters, IEEE
are more accurate than the parameters of the traditional Transactions on Control Systems Technology 19(4) (2011),
PID. 818–831.
[8] K. Kiguchi and T. Fukuda, Intelligent position/force controller
for industrial robot manipulators-Application of fuzzy neural
7. Conclusion networks, IEEE Transactions on industrial electronics 44(6)
(1997), 753–761.
[9] M.A. Ahmad, M.Z.M. Tumari and A.N.K. Nasir, Composite
This paper presents a fuzzy PID control algo- Fuzzy logic control approach to a flexible joint manipula-
rithm. The controller parameters are updated online at tor, International Journal of Advanced Robotic Systems 10(1)
each sampling time; traditional PID parameters can- (2013), 1–9.
not update online. The fuzzy PID algorithm improves [10] N. Pedrocchi, E. Villagrossi, F. Vicentini and L.M. Tosatti,
Robot-dynamic calibration improvement by local identi-
efficiency and accuracy. This proposed control strategy fication, IEEE International Conference on Robotics &
is easy to implement. The simulation results indicate Automation, 2014, pp. 5990–5997.
that the fuzzy PID algorithm meets the demands of [11] O. Lambercy, J.C. Metzger, M. Santello and R. Gassert, A
method to study precision grip control in viscoelastic force
deburring processing. Compared to the traditional PID
fields using a robotic gripper, IEEE Transactions on Biomedi-
controller, the fuzzy PID controller guarantees the sta- cal Engineering 62(1) (2014), 39–48.
bility of the closed-loop system and chooses the proper [12] P. Hamelin, P. Bigras, J. Beaudry, P.L. Richard and M. Blain,
parameters according to the control rules. The uncer- Multiobjective optimization of an observer-based controller:
Theory and experiments on an underwater grinding robot,
tainties are detected online, therefore allowing adaptive IEEE Transactions on Control Systems Technology 22(5)
compensation for error. (2014), 1875–1882.
[13] S.G. Tzafestas and G.G. Rigatos, Fuzzy reinforcement learning
control for compliance tasks of robotic manipulators, IEEE
Acknowledgments Transactions on Systems 32(1) (2002), 107–113.
[14] S. Alatartsev and F. Ortmeier, Improving the sequence of
This work was supported by the Youth Science robotic tasks with freedom of execution, IEEE International
Conference on Intelligent Robots and Systems, 2014, pp.
Fund Project NO.61305116 of the National Natural 4503–4510.
Science Foundation, and Major Science and Tech- [15] W. Li and Y. Hori, Vibration suppression using single neuron-
nology Projects of Chongqing Robot Industry based PI Fuzzy controller and fractional-order disturbance
observer, IEEE Transactions on Industrial Electronics 54(1)
No.cstc2013jcsf-zdzxqqX0005.
(2007), 117–126.
[16] X.L. Wang, Y. Wang and Y.N. Xue, Intelligent compliance con-
trol for robotic deburring using fuzzy logic, IEEE International
References Conference on Industrial Informatics, 2006, pp. 91–96.
[17] X.H. Liu, X.H. Chen, X.H. Zheng, S.P. Li and Z.B. Wang,
[1] A. Robertsson, T. Olsson, R. Johansson, A. Blomdell, Development of a GA-fuzzy-immune PID controller with
K. Nilsson and M. Haage, Implementation of industrial robot incomplete derivation for robot dexterous hand, The Scientific
force control case study: High power stub grinding and debur- World Journal (2014), 1–10.
 Y. Tao et al. / Fuzzy PID control method of deburring industrial robots 2455

[18] Y.X. Song, H.B. Lv and Z.H. Yang, An Adaptive modeling [20] Y. Zhao, J. Zhao, L.Q. Zhang, Z. Li and Q. Tang, Path planning
method for a robot belt grinding process, IEEE Transactions for automatic robotic blade grinding, IEEE International Con-
on Mechatronics 17(2) (2012), 309–317. ference on Industrial Mechatronics and Automation, 2009, pp.
[19] Y.X. Song, H.J. Yang and H.B. Lv, Intelligent control for 1556–1560.
a robot belt grinding system, IEEE Transactions on Control
Systems Technology 21(3) (2013), 716–724.