Autonomous Path Planning and Navigation of a Mobile Robot with Multi-Sensors based on Fuzzy Logic in Dynamic Environment

Yupei Yan & Yangmin Li
International Journal of Intelligent Systems and Applications in Robotics (IJRA), Volume (7):Issue (1):2016 1
Autonomous Path Planning and Navigation of a Mobile Robot
with Multi-Sensors based on Fuzzy Logic in Dynamic
Environment
Yupei Yan yb47409@umac.mo
Department of Electromechanical Engineering
University of Macau
Macao SAR, 999078, China
Yangmin Li ymli@umac.mo
Department of Electromechanical Engineering
University of Macau
Macao SAR, 999078, China
Abstract
The mobile robot is applied widely and investigated deeply in industrial fields, meanwhile, mobile
robot autonomous path planning and navigation algorithm is a hot research topic. In this paper,
firstly mobile robot is introduced, the general path planning and navigation algorithms of the
mobile robot are reviewed, then a fuzzy logic with filter smoothing is proposed based on the data
from the laser scan sensor and GPS module, which is useful for mobile robot to find the best path
to the destination automatically according to the position and size of the gaps between the
obstacles in the dynamic environment, finally our designed mobile robot and corresponding
Android APP are introduced, the path planning and navigation algorithms are tested on this
mobile robot, the testing result shows that this algorithm is globally optimized, quickly responded,
battery power and hardware cost saved compared with other algorithms, it is suitable for the
mobile robot that is running on the embedded system and it can satisfy our design requirement.
Keywords: Mobile Robot, Path Planning, Navigation, Fuzzy Logic, Embedded System.
1. INTRODUCTION
Nowadays there are extensively and widely applications of the mobile robot in various fields, such
as moving, testing and performing the scientific experiments on the Mars, serving and running as
a servant in a restaurant or house, transporting or moving the heavy goods to the designated
destinations in the warehouse etc.
The mobile robot should be designed to be more sensitive, intelligent and networked as it plays
an important role in more and more crucial occasions, it can judge and learn the operation
according to the multisensory fusion algorithm, meanwhile, it can complete and take place of the
humans’ work with high efficiency through its human-like properties [1].
As an intelligent standard of the mobile robot, path planning and navigation algorithms become
hot and critical research fields for the mobile robot [2], in order to reach the designated
destination while avoiding the obstacles automatically in the dynamic environment, many path
planning and navigation algorithms were proposed and investigated to get the best path to
destination like humans moving in a real dynamic environment, the path planning and navigation
algorithms should be designed according to the following standards: responding quickly, shorter
path distance, less time consumption, optimality, completeness, reasonability, meeting the
constraints.
The early path planning algorithm is to try finding a best path in a static environment that the

locations of the obstacles are fixed, in which A* algorithm is an assessment system of the
resource cost from the initial point to the designated point. Visibility graph algorithm can find the
best path according to the line that linked the top points of all the obstacles. With the rapid
development of path planning algorithms, the path planning algorithm that is suitable for running
in the dynamic environment is put forward, such as the artificial potential field algorithm that can
optimize the path in the partial scope, but it is not effective for the global situation, so the generic
algorithm arises which can optimize the path in the global area, but it needs high standard
hardware to calculate and the hardware cost is increased [3].
Also many navigation algorithms of the mobile robot are proposed along with the development of
the mobile robot, such as the generic algorithm which can collect the information of the
surrounding environment and judge the next step operation. Neural network technology can
accumulate the knowledge of the environment through the training of the self-organizing neural
network. The behavior algorithm will divide the navigation function into many little function
modules which have their own navigational judgments. The above algorithms are usually running
on a computer to make sure the enough processing efficiency and fast responding [4]-[5].
Because the computer is power consumption, expensive, big and heavy for the mobile robot, so
most of the mobile robot is running on an embedded system platform to save battery power and
cost, as well as easy to carry and move, therefore, in order to design a better path planning and
navigation algorithm that is hardware cost reduced, battery power saved, responding quickly and
globally optimized for the mobile robot, the fuzzy algorithm is proposed because of quickly
responding and less calculation, it is suitable for the mobile robot that is running on the platform
of embedded system in a dynamic environment.
According to the above analysis, we design and manufacture a mobile robot with the platform of
“ARM” CPU and embedded Linux system, running path planning and navigation algorithms with
fuzzy logic and filter smoothing processing. Android APP and video monitoring with internet
browser function are added in the system. This mobile robot is tested and the testing result of
path planning and navigation algorithm can meet the design requirement.
2. PATH PLANNING AND NAVIGATION
2.1 Multi-Sensor Data
The laser sensor is used to measure the angle, distance, signal quality, synchronized bit of each
obstacle within 6 meters, it is produced by Rplidar company of China, the data of obstacles is
output from laser sensor per round as Figure 1 shows.
FIGURE 1: Laser Sensor Scanning Data Graphic.
The Figure 2 shows the corresponding simulated environment, in which the measured line OC is
the longest distance, the line AB is the gap between the two obstacles.
It is necessary to get the accurate position of the mobile robot to see whether it has reached the

destination, so the GPS sensor is needed to output the real-time longitude and latitude, which is
helpful for the mobile robot to reach destination after comparing with the designated longitude
and latitude. The mobile robot should also detect its moving direction which helps the mobile
robot to correct its moving direction in order to get closer to the destination [6].
FIGURE 2: Simulated environment according to the laser sensor scanning data.
In order to make sure that the position and direction judgement are accurate, the flight control
module is added in the system which integrates the GPS sensor, gyroscope, accelerometer,
magnetometer and electronic compass sensors together, it is “HT-Hawk” module which is
produced by “HengTuo” company, the integrated GPS sensor is the new version UBLOX GPS
module NEO-M8N, the outputted longitude and latitude can reach the accuracy of 0.9 meters.
2.2 Path Planning
The mobile robot’s width is about 580mm and the length is about 800mm, this is a constraint
condition to process the laser sensor data, so it is necessary to draw the graphic of the obstacles
and gaps surrounded for further analysis.
FIGURE 3: Laser Sensor Scanning with Gaps.
The laser sensor will process it as a gap if the measured distance is larger than 6 meters, in
Figure 3, the gap size AB can be calculated by law of cosines:
2 2
2 cos 1AB OA OB OA OB       (1)
For the scanning data in the gap of AB, such as the K point, they are
0) Kdistance(O , 0)( OKlitysignal_qua (2)

The distance data on the obstacles are all positive values, such as the distance data of points on
line HA, so the points of A, B can be confirmed, the distance OA, OB and the angles between
point A and B can be attained
1 B A   
(3)
The other distance and angles can be acquired in the same way, such as distance of OC, OD,
OE, OF, OG, OH, gaps size of CD, EF, GH, angle of 1 , 3 , 4 , it should be noted that the gap
GH across the 0 , the 4 can be calculated by
4 360 -H G    
(4)
For the consideration of the mobile robot’s width and the corresponding redundancy, the mobile
robot will choose the gap size which is larger than 600mm to compare each other to find the max
gap size, in Figure 3, the max gap size in one round is EF, so the mobile robot should run to the
direction of gap EF, it is suitable to move to the angle of the middle position of gap EF and it is:
  3
( )
2
angle OI angle OE

 
(5)
It is worth noted that if there is an obstacle in front of the mobile robot, such as the red “Obstacle
M” in Figure 3, the mobile robot can’t move forward and the constraint condition is:
If the laser sensor scanning angle is in 320 ~ 40 , checking:
if mmdistance 1000 , the variable 1rdstop_forwa 
else 0rdstop_forwa
It is the same way to process the situation for moving backward of the mobile robot:
When the laser sensor scanning angle is in 160 ~ 200 , checking:
if mmdistance 1000 , the variable back 1stop_ ward 
else 0back wardstop_
The )(OIangle in formula (5) is the destination angle, the following Table 1 shows the relationship
between destination angle and corresponding turning direction of the mobile robot, if
1stop_forward  or back 1stop_ ward  , the mobile robot will turn right as the default direction.
Destination angle Turning
direction
340 ~ 20 while 0rdstop_forwa forward
340 ~ 20 while 1stop_forward  right
20 ~ 160 right
160 ~ 200 while 0back wardstop_ backward
160 ~ 200 while back 1stop_ ward  right
200 ~ 340 left
TABLE 1: Relationship of destination angle and turning direction.
2.3 Navigation
Based on the above analysis, the largest gap can be located and the mobile robot can turn the
direction to go through the largest gap, but it is meaningless because this only provides the

widest path for mobile robot to move ahead while avoiding the obstacles on the path, in order to
reach the destination to fulfill some works for the mobile robot in the real world, it is necessary to
designate the destination and mobile robot can arrive the destination in dynamic environment, the
destination position can be located through designating the longitude and latitude on the Android
APP, the mobile robot should navigate and plan the path according to the sensors fusion
algorithms [7]-[8].
FIGURE 4: The navigation without rotation of the mobile robot.
As Figure 4 shows, compared the designated and detected longitude and latitude, it can be
confirmed that the destination is on the northeast position of the mobile robot, according to the
above path planning algorithm, the mobile robot will find the max gap in the scope between OC
and OD to judge the max gap direction, if there is no rotation of the mobile robot, the forward
direction of the mobile robot points to 0 which is just the north direction, in this case
  0angle OC  
(6)
  90angle OD  
(7)
The mobile robot will find the max gap between 0 to 90 which is towards to the northeast
direction of the mobile robot, in Figure 4, there are two gaps between 0 to 90: d1 and d2,
because d1 is less than the width of the mobile robot and d1 cannot be chosen, d2 satisfy the
condition and the max gap between 0 to 90 is d2, so the mobile robot will turn right to rotate
3 and will move towards to the OB direction.
But in most instances, the mobile robot will rotate to some degrees and the forward direction is
not towards to the north, in Figure 5, the mobile robot rotates 1 and its forward direction is
towards to OA now, because the destination is still on the northeast position of the mobile robot, it
will find the max gap between OC and OD, but the OC is not 0 and OD is not 90 for mobile
robot now, so it is necessary to do some degree transformation in this situation.

  0angle OD 
2 0 
  0angle OC 
FIGURE 5: The navigation with rotation of the mobile robot.
In order to obtain the max gap between the scope of OC and OD, it is necessary to calculate the
degree of OC and OD in this situation, in Figure 5, the mobile robot has been rotated 1 with
respect to north direction, the reference system will also rotate 1 to meet transformation of
coordinates, so all the degrees will reduce 1 to make the new coordinate effective.
  1angle OC  
(8)
  190angle OD   
(9)
2 3 1   
(10)
If the destination orientation is not on the northeast position after comparison of the designated
and detected longitude and latitude, such as the mobile robot is on the southeast, southwest and
northwest position of the destination, the  angle OC and  angle OD are different and they can
be listed in the Table 2.
Destination
orientation
Degree scope of finding max gap
Northeast   1angle OC     190angle OD  
Southeast   190angle OC     1180angle OD  
Southwest   1180angle OC     1270angle OD  
Northwest   1270angle OC     1360angle OD  
TABLE 2: Relationship between destination orientation and degree scope.
If there is minus degree, it needs the correction:
If ,   1360angle OC   (11)
if ,   190 360angle OD     (12)
if , 2 3 1 360      (13)
In Figure 5, mobile robot scans the gaps between OC and OD and attaches the max gap d2, the
mobile robot should turn left according to Table 1 to go towards to d2, scanning obstacles and
attaining gaps need the help of the laser scan sensor, the comparison of the longitude and

latitude is fulfilled with the help of GPS sensor, detecting the rotation direction of the mobile robot
is completed through the data from gyroscope, accelerometer, magnetometer and electronic
compass sensors [9].
There is another situation that there is no gap satisfying the condition between the OC and OD,
just as d1 and d2 in the Figure 6 shows, both of d1 and d2 are less than 600mm and the mobile
robot can’t go through the space from OC to OD, so the laser scan sensor will find the max gap in
other space to get the satisfied gap, as the Figure 6 shows, OA is the forward direction of the
mobile robot, and the mobile robot finds two satisfied gaps which is d3 and d4, according to the
above path planning algorithm, d4 is the largest gap and it should go to the d4 towards to the OE
direction, but for the navigation, it not only focuses the gap size, more important factor is the gap
direction which is crucial for the mobile robot to get closer to the destination, after transformation
of the coordinate, it finds that the d3 is closer to the destination, so the mobile robot will turn right
according to Table 1 to go through d3 and move towards to the OB direction.
FIGURE 6: The navigation of mobile robot without satisfied gap.
If there is no satisfied gap whose length is more than 600mm in the whole round, the mobile robot
will stay static and not move, until there is a satisfied gap available again and it will move
according to the above algorithm to reach the destination.
2.4 Fuzzy Logic
From above analyses, it is clear that fuzzy logic is suitable to process the path planning and
navigation of the mobile robot, which makes the turning direction of the mobile robot more stable
and reliable [10]-[11].
FIGURE 7: Fuzzy logic for path planning and navigation of the mobile robot

The Figure 7 shows the fuzzy logic for the path planning and navigation of the mobile robot, the
input variables are the measured direction and angle of the scanned gaps, as well as the moving
direction, longitude and latitude of the mobile robot, the membership functions can be listed as
follows.















340200
360340
340360
340
20
20
a
a
a
a
a
dforward (14)
20
20 90
160 90
160
90 160
90 20
0 20 160
right
a
a
a
ad
a or a
















 


 

 
(15)
160
160 180
200 180
200
180 200
180 160
0 160 200
backward
a
a
a
ad
a or a

  

   


  
(16)
200
200 270
340 270
340
270 340
270 200
0 200 340
left
a
a
a
ad
a or a

  

   


  
(17)
FIGURE 8: The membership function of fuzzy logic.
The mobile robot rotates according to the turning direction until it reaches the forward or
backward direction, this process repeats continuously for the mobile robot till it moves to the
destination [12]-[13].
The application of fuzzy logic can save the CPU calculation and respond quickly, which is suitable
for the mobile robot that is designed on a platform of the embedded system.
2.5 Filter Smoothing
In the test, the mobile robot changes moving direction very quickly sometimes, about 2 to 3 times
per second, so the mobile robot can’t move stably on the attained path, after the debug
information comes from the serial port, there are some angle noises data inducing the incorrect
turning direction in Table 1, because the laser sensor scanning work is running in one thread, the
time slice task management in the Linux system will induce the time of entering the laser sensor
scanning thread is uncertain, it is possible that the laser sensor only scans a part of the space
and gets the incomplete data when it enters in the thread, so it is necessary to count the turning
direction result per second to make the turning direction data more smoothly [14]-[15].

FIGURE 9: Noise angle and normal angle distribution in each second.
In order to make sure the mobile robot to turn the direction smoothly, the noise angle should be
deleted through the statistics of angles’ number in different turning direction per second, in Figure
9, there are total 7 angles, according to Table 1, the number of angles in right direction scope is 5,
which is more than the number of other directions, so the other 2 angles are the noise angles with
the red circles, the mobile robot will move right in this second, this process continues per second
to make the mobile robot move more stably and accurately.
3. TEST RESULT
In order to test the path planning and navigation more practically in the real dynamic environment,
we have designed a mobile robot running on the embedded platform with “ARM” CPU and
embedded Linux system as Figure 10 shows.
In order to intensify the moving driven power of the mobile robot, the DC motor connects with a
gear motor with the reduction ratio 1:50 to increase the turning torque, the belt connects two
wheels together on each side and there are total 4 wheels on this mobile robot. According to the
testing results, the mobile robot can move properly when there is a man whose weight is 86KG
standing on the mobile robot, so this mobile robot can carry the heavy goods in the real life
applications.
FIGURE 10: The designed mobile robot.
There are three parts of the hardware platform for the mobile robot, the first part is ARM
development board with “S5PV210” CPU which is running the embedded Linux system, it can
realize the function of the camera video display, LCD display, operation on touch screen, laser
scan sensor data collection, data fusion of GPS and other sensors, bluetooth and wifi data
transmission, path planning and navigation algorithm.

The second part is another MCU board that connects to the ARM board through serial port, which
could transmit the motor control command and the feedback speed of motors with the designed
protocol, because there is calling and task schedule delay for the Linux system, this MCU board
is designed to read the 25KHz motor speed feedback pulse accurately.
The third part is the “SCM6716” board to supply the large current and power to drive the motor
running normally, the peak current can reach about 1.8A and this will make sure the mobile robot
can bear and carry the heavy objects in the real life applications.
FIGURE 11: The LCD and touchscreen with camera video display that is running in the embedded Linux
system.
The other hardware contains LCD and touch screen, the 2.0M camera, Serial-bluetooth module,
USB-wifi module, laser scan sensor, GPS sensor, gyroscope, accelerometer, magnetometer and
electronic compass sensors.
The GUI utilizes QT graphic library to show the widgets on the LCD as shown in Figure 11,
through calling the corresponding signal and slot function, the QT API fulfills the LCD and touch
screen programming work in the Linux developing environment.
FIGURE 12: Android APP.
The Android APP is designed to control the mobile robot more easily and freely, it communicates

with the ARM board with Serial-bluetooth module to control the mobile robot running status,
working mode, moving direction and speed, designated longitude and latitude etc.
FIGURE 13: Remote monitoring with internet browser.
The mobile robot should be more networked along with the network technology developing, the
mjpeg-stream is used to transmit the camera video data to the internet browser cross various
platforms with the TCP/IP protocol, it can remotely monitor on the Firefox, browser it is
transplanted in the embedded Linux system of the mobile robot, the testing result shows it can
work properly and it is helpful to know about the status of mobile robot immediately via the
internet browser.
FIGURE 14: Linux programming software logic structure of the mobile robot.
The GUI framework is the Linux programming software entrance, it points to the QT interface
pointer, after transmitting and displaying the camera data to the QT interface, it creates three
threads to finish the work of motor controlling, laser sensor scanning, the path planning and
navigation algorithm, refresh the widget content [16].

FIGURE 15: Matlab simulation for navigation of mobile robot with stratified gap.
Figure 15 shows the Matlab simulation for the navigation of the mobile robot when there is
satisfied gap between two obstacles, the green circle is the mobile robot, the black object is the
obstacles, the red rectangle is the destination, the mobile robot can detect the direction and
distance of the gap between the obstacles through the laser scan data, if there is a gap that is
wider than the width of the mobile robot, it could go through the gap and reach the destination.
FIGURE 16: Matlab simulation for navigation of mobile robot without stratified gap
In the Figure 16, the gap between two obstacles is too narrow and there is no stratified gap
between them, the mobile robot should find other ways to bypass the obstacle as shown in the
Figure 16, this operation continues until there is a satisfied gap detected on the path, and it will go
to the destination finally. Optimal path planning will be investigated via PSO algorithm [17]-[19].
4. CONCLUSION
The testing result shows that the proposed path planning and navigation algorithm based on
fuzzy logic and filter smoothing are effective and real-time responding, compared with other
algorithms, it is simply computed and realized for the mobile robot designed on the platform of the
embedded system that is running in the real dynamic environment.
In the designing process, one of the most challenges is showing the real-time camera video on
the LCD screen, because the video is stream data in Linux system and the data is too big to be

processed immediately. The other challenge is to get rid of the noise data from the laser sensor, it
is difficult to analyze and find the data errors when reading a lot of scanned data in a short time,
but the final testing results show that our designed mobile robot can run normally with the
designed path planning and navigation algorithm.
The advantage of this algorithm is easy to realize, respond quickly, save reacting time, save CPU
calculations, further save power supply comparing with other algorithms, they are especially
important for the embedded system with the limited CPU calculation and battery power, this can
make sure the mobile robot process the sensors fusion data and find the best path to navigate
quickly and accurately, the testing results show that it can meet our design requirements.
5. ACKNOWLEDGEMENT
This work was supported from Macao Science and Technology Development Fund (108/ 2012/A3,
110/2013/A3), Research Committee of University of Macau (MYRG2015-00194-FST).
Lastly Yupei Yan would like to thank to his laboratory’s colleagues Xiao Xiao, Yiru Tang, Gen
Zhang, Zhigang Wu, Song Lu and Ping He etc., they help to construct the mechanical part of the
mobile robot or give many suggestions on how to write and improve the quality of this paper.
6. REFERENCES
[1] I. Hassanzadeh, K. Madani , M. A. Badamchizadeh, “Mobile robot path planning based on
shuffled frog leaping optimization algorithm,” IEEE International Conference on Automation
Science and Engineering, Toronto, pp.680 – 685, August 21-24, 2010.
[2] Y. M. Li and Y. G. Liu, “Real-time tip-over prevention and path following control for redundant
nonholonomic mobile modular manipulators via fuzzy and neural-fuzzy approaches,” Journal
of Dynamic System, Measurement, and Control, Transactions of ASME, vol.128, no.4,
pp.753-764, December, 2006.
[3] B. Park, W. K. Chung, “Efficient environment representation for mobile robot path planning
using CVT-PRM with Halton sampling,” Electronics Letters, pp. 1397 – 1399, October, 2012.
[4] Z. Huang, J. Zhu, L. Yang, B. Xue, “Accurate 3-D position and orientation method for indoor
mobile robot navigation based on photoelectric scanning,” IEEE Transactions on
Instrumentation and Measurement, pp.2518 – 2529, April, 2015.
[5] C. F. Juang, Y. C. Chang, “Evolutionary-group-based particle-swarm-optimized fuzzy
controller with application to mobile-robot navigation in unknown environments,” IEEE
Transactions on Fuzzy Systems, pp. 379-392, January, 2011.
[6] Y. M. Li and Y. G. Liu, “Fuzzy logic self-motion planning and robust adaptive control for tip-over
avoidance of redundant mobile modular manipulators,” IEEE/ASME International Conference
on Advanced Intelligent Mechatronics(AIM), Monterey, California, USA, pp.1281-1286, 24-28
July, 2005.
[7] K. K. Zhao and Y. M. Li, “Path planning in large-scale indoor environment using RRT,” 32nd
Chinese Control Conference (CCC), Xian, China, pp.5993-5998, July 26-28, 2013.
[8] Y. Zhang, D.W. Gong, and J.H Zhang, “Robot path planning in uncertain environment using
multi-objective particle swarm optimization,” Neurocomputing, vo1.103, pp.172-185, 2013.
[9] E. Keshavarz and E. Khorram, “A fuzzy shortest path with the highest reliability,” Journal of
Computational and Applied Mathematics, vol. 230, pp.204-212, 2009.

[10] Y. M. Li and Y. G. Liu, “Online fuzzy logic control for tipover avoidance of autonomous
redundant mobile manipulators,” International Journal of Vehicle Autonomous Systems, vol.4,
no.1, pp.24-43, 2006.
[11] Y. G. Liu and Y. M. Li, “Dynamic modeling and adaptive neural-fuzzy control for nonholonomic
mobile manipulators moving on a slope,” International Journal of Control, Automation, and
Systems, vol.4, no. 2, pp. 197-203, 2006.
[12]Y. M. Li and Y. G. Liu, “Obstacle avoidance for redundant nonholonomic mobile modular
manipulators via neural fuzzy approaches,” Advances in Natural Computation-ICNC, Eds by L.
Wang, K. Chen and Y.S. Ong, Springer, LNCS 3612, pp. 1109-1118, 2005.
[13]Y. M. Li and Y. G. Liu, “Robust adaptive neural fuzzy control for autonomous redundant
nonholonomic mobile modular manipulators,” International Journal of Vehicle Autonomous
Systems, vol.4, nos.2-4, pp.268-284, 2006.
[14]X. Chen and Y. M. Li, “Smooth formation navigation of multiple mobile robots for avoiding
moving obstacles,” International Journal of Control, Automation, and Systems, vol. 4, no.4,
pp.466-479, 2006.
[15]J.-G. Wang and Y. M. Li, “Dynamic modeling of the mobile humanoid robot,” IEEE
International Conference on Robotics and Biomimetics (ROBIO), Bangkok, Thailand, pp.639-
644, February 22-25, 2009.
[16]Y.C. Song, Y. M. Li and X. Xiao, “Control system design and study for an automatic mobile
robot,” IEEE International Conference on Information and Automation (ICIA), Lijiang, Yunnan,
China, pp. 3118 – 3123, August 8 - 10, 2015.
[17]X. Chen and Y. M. Li, “A modified PSO structure resulting in high exploration ability with
convergence guaranteed,” IEEE Transactions on System, Man and Cybernetics: Part B,
vol.37, no.5, pp.1271-1289, October 2007.
[18]G. Zhang and Y. M. Li, “Parallel and cooperative particle swarm optimizer for multimodal
problems,” Mathematical Problems in Engineering, Article ID 743671, 10 pages, 2015.
[19]G. Zhang and Y. M. Li, “A memetic algorithm for global optimization of multimodal
nonseparable problems,” IEEE Transactions on Cybernetics, vol. 46, no.6, pp. 1375-1387,
June 2016.

Autonomous Path Planning and Navigation of a Mobile Robot with Multi-Sensors based on Fuzzy Logic in Dynamic Environment

Recommended

More Related Content

What's hot (18)

Similar to Autonomous Path Planning and Navigation of a Mobile Robot with Multi-Sensors based on Fuzzy Logic in Dynamic Environment (20)

Recently uploaded (20)

Autonomous Path Planning and Navigation of a Mobile Robot with Multi-Sensors based on Fuzzy Logic in Dynamic Environment