An Improved Ultrasonic-Based Localization Using ... - Semantic Scholar
2009 International Asia Conference on Informatics in Control, Automation and Robotics
An Improved Ultrasonic-Based Localization Using Reflection Method Chen-Chien Hsu
Hsin-Chuan Chen
Chien-Yu Lai
Department of Electrical Engineering Tamkang University Tamsui, Taipei, Taiwan e-mail: [email protected]
Department of Electronic Engineering St. John’s University Tamsui, Taipei, Taiwan e-mail: [email protected]
Department of Electronic Engineering St. John’s University Tamsui, Taipei, Taiwan e-mail: [email protected]
on techniques of indoor localization. It is therefore the objective of this paper to propose an ultrasonic-based localization using ultrasonic reflection approach to collect distance measurements for calculating the coordinate of a mobile robot based on an established triangular relationship, circumventing complex mathematical derivations [13].
Abstract—In this paper, an ultrasonic-based localization using ultrasonic reflection method is proposed. Experiment environment includes a mobile robot and four poles forming a square around the measuring site. Ultrasonic sensors built on the poles serve as receivers and the mobile robot serves as a transmitter, and all ultrasonic sensors are integrated with their Zig-Bee modules. By the sequential ultrasonic signal transmission between the robot and the poles, the ultrasonic sensors on the poles can then measure the time-of-flight (TOF) without interference to calculate the distance between the receiver and transmitter ends. According to an established twodimensional coordinate model, position of the robot can be obtained based on the distance measurements. Thanks to Zig-Bee modules, position information of the robot can be instantly conveyed to a remote PC for monitoring the robot and path planning. Extensive experiments conducted have shown a satisfactory accuracy of the coordinates of the mobile robot can be obtained via the proposed localization scheme.
II.
DERIVATION OF 2D COORDINATES FOR ULTRASONICBASED LOCALIZATION
Figure 1 illustrates a simplified localization environment for deriving 2D coordinates of a target object P(X,Y) with a ultrasonic transmitter U0 by using four ultrasonic sensors U1~U4 as receivers. Assume distance between U1 and U2 is B and distance between U1 and U4 is A, and Z1~Z3 are the distances from point P measured by ultrasonic sensors U1-U3, respectively. Based on well known triangular relationships, we have distance measurements between the object and the ultrasonic sensors U1, U2, and U3 as:
Keywords-ultrasonic reflection; localization; robot; Zig-Bee; 2D coordinates
2
I.
Z1 = X 2 + Y 2
(1)
Z 2 = ( B − X) 2 + Y 2
2
(2)
Z 3 = (B − X) 2 + ( A − Y ) 2
(3)
INTRODUCTION
There exist several localization methods; Global Positioning System (GPS) [5] is the most popular one, which has been widely applied with versatile applications [6]-[9]. Unfortunately, signals are significantly blocked out when GPS is used indoor. As a result, GPS is not suitable for indoor environment in general. Various indoor localization techniques are therefore developed in recent years, including RFID [10], Zig-Bee [4], and wireless networks, etc [11], based on signal strength. These methods, however, suffer from inaccurate measurements because the signal strength is generally deteriorated by various noises in the environment. Alternatively, ultrasonic sensor among many different types of available sensors is being extensively used since its measurement principle is simple and thus the hardware implementation is easy. Particularly, the accuracy of ultrasonic sensors has reached the dimension of centimeter [12]. It is critically important for a mobile robot to move from place to place without running into obstacles or getting lost. Successful management of this navigation task depends on a robust and reliable solution to the estimation of real robot position in the world space [14]. Current trend toward selflocalization of mobile robots is therefore extensively studied [1]-[3]. Numerous practical industrial applications, for example real-time moving object monitoring [4], are based
978-0-7695-3519-7/09 $25.00 © 2009 IEEE DOI 10.1109/CAR.2009.93
2
Combining (1) and (2), we have 2
2
Z1 − X 2 = Z 2 − B2 + 2BX − X 2
(4)
Therefore, the x-coordinate of the object P can be given by: 2
2
Z − Z2 + B2 X= 1 2B
(5)
Similarly, combining (2) and (3), we have 2
2
Z 3 = Z 2 − Y 2 + ( A − Y) 2
(6)
From (6), we can obtain the y-coordinate of the object P: 2
Y=
2
Z 2 − Z3 + A 2 2A
437
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
(7)
A. System Architecture Figure 2 shows the proposed ultrasonic-based localization scheme, where the target object is a mobile robot incorporating an ultrasonic sensor U0 as a transmitter. A micro-controller, H8tiny-3694 (Renesas Technology Corp.), is embedded into the robot for activating ultrasonic transmitter and communicating with its Zig-Bee module. Experiment environment also includes four poles forming a square around the measuring site, where an ultrasonic sensor is built on the top of each pole as a receiver. Based on ZigBee modules integrated with the ultrasonic sensors, a wireless network can be established for seamless communication among the ultrasonic sensors. As a result, the main controller PC can promptly give instructions to the ultrasonic sensors to transmit and receive signals. Distance measurements collected by 4 receivers of the measuring site are processed by the remote PC under an established 2D coordinate model to estimate the position of the robot.
U3
U4 Z3 P(X,Y)
U0 A Z1
Z2
Y
U1
U2
X
B
B. Ultrasonic Reflection Due to the angle limitation of ultrasonic signal, the detection capability will be affected, and it further limits the measurable area. In this paper, we try using a cone with 45° incline to reflect the incident ultrasonic signal from the ultrasonic transmitter to form a 360° plane, such that 4 ultrasonic sensors as receivers can easily receive the transmitting ultrasonic signal everywhere the mobile object moves in a detection area. Moreover, it is unnecessary for the ultrasonic transmitter to adjust its transmitting direction when one ultrasonic receiving sensor switches to the other one. Figure 3 illustrates how the transmitting ultrasonic signal forms a 360° ultrasonic plane via cone’s reflection.
Figure 1. A simplified localization scheme using ultrasonic-based method.
Comparing the measurement results obtained with predefined conditions, we can evaluate the feasibility of the coordinates if abnormal measurements occur. Thanks to these procedures for screening the measurements, accuracy of the coordinate of the object P can be significantly improved. III.
IMPROVED ULTRASONIC-BASED LOCALIZATION SYSTEM
To precisely estimate the position of a target object, for example, the mobile robot, distance measurements need to be obtained via the ultrasonic sensors. Subsequently, the distance measurements are used to calculate the coordinate of the robot based on an established two-dimensional (2D) coordinate model via a triangular relationship described as Section 2. In this system, in addition to ultrasonic sensors, Zig-Bee wireless modules are also required to transmit the related position information to the main controller PC. Furthermore, using ultrasonic reflection method can improve the localization performance due to solving the angel limitation of ultrasonic signal. U4
C. Operations At initial state, the main controller PC issues a command to 4 ultrasonic sensors at the receiving ends to prepare for receiving signals. At the same time, a command is issued to a specific ultrasonic sensor at the transmitting end on the robot via the Zig-Bee wireless network for transmitting ultrasonic signal. Once the ultrasonic sensors at the receiving ends receive signals, conversion into distance measurements is performed so that distance measurement can be conveyed to the main controller PC for further processing. Based on the collected distance measurements, position of the robot can then be calculated according to a 2D coordinate model.
U3
Ultrasonic sensor as transmitter
U0
Remote PC
U2
U1 : Ultrasonic signal
: Zig-Bee signal
: Ultrasonic sensor
: Ultrasonic sensor
as receiver
Cone for refraction : Incident wave : Reflection wave
as transmitter
Figure 3. Ultrasonic reflection for an ultrasonic transmitter.
Figure 2. Proposed ultrasonic-based localization scheme.
438
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
D. Zig-Bee Wireless Network The Zig-Bee module for each ultrasonic sensor, ZIG100B, provides three communication modes, including Peer to Peer, Waiting, and Broadcasting. In this paper, the Broadcasting mode is adopted because each receiver on the pole passively receives commands from the remote PC, which is suitable for broadcasting mode. Its simplicity for configuration and implementation is also an advantage.
Start
Output the localization results
Initialize (n=1)
Calculate position of the robot via a 2D coordinate m odel
nth ultrasonic sensor at the receiving end waits for receiving signals
IV.
Yes Polling com pleted? (n=4?)
Activate ultrasonic transmitter to transm it signals
Proceed with next ultrasonic sensor n=n+1
nth ultrasonic sensor at the receiving end receives and stores distance measurements
No
In our experiments, 9 points uniformly distributed in the measuring environment are sampled for testing. Experiment results based on the proposed ultrasonic-based localization are listed in Table 1. We find that the mean errors of the measurements lie within an acceptable range of ± 10cm, and the entire measurement area also can be enlarged to 300cm2 by reflecting the transmitting ultrasonic signal. Note that distance measurements are collected by way of polling in this paper so as to avoid interference that might have occurred. As a result, each receiver is activated to receive signal one at a time in a sequential order. Data collection efficiency is inevitably slowed down because of the polling process. For partially solving this problem, the independent remote PC in this proposed localization scheme is responsible for constantly consolidating distance measurements to calculate 2D coordinates of the robot.
Figure 4. Operations of the ultrasonic-based localization system. TABLE I.
Point Number
EXPERIMENT RESULTS OF PROPOSED LOCALIZATION SYSTEM
P(X,Y) (cm) Real Measured Coordinate Coordinate
Distance Error (cm)
1
(100,100)
(101,99)
(1,1)
2
(200,100)
(196,101)
(4,1)
3
(300,100)
(295,103)
(5,3)
4
(100,200)
(100,198)
(0,2)
5
(200,200)
(198,197)
(2,3)
6
(300,200)
(293,198)
(7,2)
7
(100,300)
(101,296)
(1,4)
8
(200,300)
(198,294)
(2,6)
9
(300,300)
(296,294)
(4,6)
EXPERIMENT RESULTS
V.
CONCLUSIONS
In this paper, an ultrasonic-based localization system for mobile robot is proposed to achieve a satisfactory accuracy in position, and solve the angle limitation of ultrasonic signal by using ultrasonic reflection. To provide seamless communication while avoiding interference in the measuring environment, Zig-Bee modules are integrated into all ultrasonic sensors for transmitting commands and receiving measurements by a sequential control. Based on an established 2D coordinate model, extensive experiments conducted have shown a satisfactory accuracy of ± 10cm in measuring the coordinates of the mobile robot can be obtained via the proposed localization scheme. When working with existing indoor localization methods, which are capable of identifying target objects in a larger area, the proposed localization scheme can further estimate the 2D coordinate of the object in better precision. Further extension of this work includes fusion of sensor data so as to provide more accurate localization results.
The ultrasonic sensors receive ultrasonic signal at the same frequency of 40k Hz. As a result, interference is inevitably encountered when all the 4 receivers receive signals at the same time. To avoid the interference, a sequential operation to control each ultrasonic sensor is employed in this work. Figure 4 shows the operation flow chart for this proposed ultrasonic-based localization system. To complete a polling cycle, ultrasonic sensors U1-U4 first enter a state of waiting to receive signal. Depending on commands received, a specific transmitter U0 is responsible for transmitting signal so that 4 receivers can detect and collect the measurements based on the time-of-flight (TOF). After a polling cycle is completed, 4 sets of distance measurements are ready for processing to calculate the coordinate of the object via a 2D coordinate model.
REFERENCES [1]
[2]
D. Tan, Q. Wang, Y. Wang, and R. Cao, “Research on position technology of indoors automobile robot,” Machinery & Electronics, vol. 9, pp. 46-47, 2005. J. Tang, T. Bao, and Z. Cai, “A localization method for mobile robot with optic mice in indoor environment,” Control & Automation, vol. 5, no. 21, pp. 20-22, 2005.
439
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
[3]
[4]
[5] [6]
[7] [8]
[9]
Y. H. Mu, Y. X. Yin, X. Y. Tu, and T. Q. Sun, “New real-time localization and motion planning method for indoor mobile robot,” Application Research of Computers, vol. 8, no. 24, pp. 106-108, 2007. K.E. Jianhua, H. Shen, and X. Wei, “Study on the ZigBee technology-based personnel position system of the underground coal mine,” Modern Electronics Technique, vol. 23, pp. 12-14, 2006. P. Enge and P. Misra, “Special issue on global positioning system,” Proceedings of the IEEE, vol. 87, no. 1, 1999. Y. Li, D. Xu, F. Li, M. Bai, and S. Zhang, “Application of GPS technology in agricultural land leveling survey,” Transactions of The Chinese Society of Agricultural Engineering, vol. 1, no. 21, pp. 6670, 2005. C. Shi, “Design of GPS vessel monitoring system,” Journal of Wuhan University of Technology, vol. 4, no. 26, pp. 448-450, 2002. Z. B. Feng, Y. S. Hua, X. Z. Qing, W. Y. Hua, and Z. Y. Zhong, “Application of GPS to the measurement of the roads,” Acta Metrologica Sinica, vol. 2, no. 23, pp. 98-105, 2002.
[10]
[11]
[12]
[13]
[14]
G. Ning, “GPS technique and its application in railway transportation engineering,” Journal of Changsha Railway University, vol. 3, no. 17, pp. 7-11, 1999. Y. Sun and P. Z. Fan, “RFID technology and its application in indoor positioning,” Computer Applications, vol. 5, no. 25, pp. 1205-1208, 2005. M. H. Zhang, S. S. Zhang and J. Cao, “Received-signal-strengthbased indoor location in wireless LANs,” Computer Science, vol. 24, no. 6, pp. 68-75, 2007. K. Zhang and G. Liu, “Research on a method of improving ultrasonic ranging precision,” Modern Electronics Technique, vol. 15, pp. 139-141, 2007. Y. L. Zhu, “Parameter research on ultrasonic orientation electrocircuit system,” Development & Innovation of Machinery & Electrical Products, vol. 19, no. 2, pp. 128-130, 2006. M. Piasecki, “Global localization for mobile robots by multiple hypothesis tracking,” Robotics and Autonomous Systems, vol. 16, no. 1, pp. 93-104, 1995.
440
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
An Improved Ultrasonic-Based Localization Using Reflection Method Chen-Chien Hsu
Hsin-Chuan Chen
Chien-Yu Lai
Department of Electrical Engineering Tamkang University Tamsui, Taipei, Taiwan e-mail: [email protected]
Department of Electronic Engineering St. John’s University Tamsui, Taipei, Taiwan e-mail: [email protected]
Department of Electronic Engineering St. John’s University Tamsui, Taipei, Taiwan e-mail: [email protected]
on techniques of indoor localization. It is therefore the objective of this paper to propose an ultrasonic-based localization using ultrasonic reflection approach to collect distance measurements for calculating the coordinate of a mobile robot based on an established triangular relationship, circumventing complex mathematical derivations [13].
Abstract—In this paper, an ultrasonic-based localization using ultrasonic reflection method is proposed. Experiment environment includes a mobile robot and four poles forming a square around the measuring site. Ultrasonic sensors built on the poles serve as receivers and the mobile robot serves as a transmitter, and all ultrasonic sensors are integrated with their Zig-Bee modules. By the sequential ultrasonic signal transmission between the robot and the poles, the ultrasonic sensors on the poles can then measure the time-of-flight (TOF) without interference to calculate the distance between the receiver and transmitter ends. According to an established twodimensional coordinate model, position of the robot can be obtained based on the distance measurements. Thanks to Zig-Bee modules, position information of the robot can be instantly conveyed to a remote PC for monitoring the robot and path planning. Extensive experiments conducted have shown a satisfactory accuracy of the coordinates of the mobile robot can be obtained via the proposed localization scheme.
II.
DERIVATION OF 2D COORDINATES FOR ULTRASONICBASED LOCALIZATION
Figure 1 illustrates a simplified localization environment for deriving 2D coordinates of a target object P(X,Y) with a ultrasonic transmitter U0 by using four ultrasonic sensors U1~U4 as receivers. Assume distance between U1 and U2 is B and distance between U1 and U4 is A, and Z1~Z3 are the distances from point P measured by ultrasonic sensors U1-U3, respectively. Based on well known triangular relationships, we have distance measurements between the object and the ultrasonic sensors U1, U2, and U3 as:
Keywords-ultrasonic reflection; localization; robot; Zig-Bee; 2D coordinates
2
I.
Z1 = X 2 + Y 2
(1)
Z 2 = ( B − X) 2 + Y 2
2
(2)
Z 3 = (B − X) 2 + ( A − Y ) 2
(3)
INTRODUCTION
There exist several localization methods; Global Positioning System (GPS) [5] is the most popular one, which has been widely applied with versatile applications [6]-[9]. Unfortunately, signals are significantly blocked out when GPS is used indoor. As a result, GPS is not suitable for indoor environment in general. Various indoor localization techniques are therefore developed in recent years, including RFID [10], Zig-Bee [4], and wireless networks, etc [11], based on signal strength. These methods, however, suffer from inaccurate measurements because the signal strength is generally deteriorated by various noises in the environment. Alternatively, ultrasonic sensor among many different types of available sensors is being extensively used since its measurement principle is simple and thus the hardware implementation is easy. Particularly, the accuracy of ultrasonic sensors has reached the dimension of centimeter [12]. It is critically important for a mobile robot to move from place to place without running into obstacles or getting lost. Successful management of this navigation task depends on a robust and reliable solution to the estimation of real robot position in the world space [14]. Current trend toward selflocalization of mobile robots is therefore extensively studied [1]-[3]. Numerous practical industrial applications, for example real-time moving object monitoring [4], are based
978-0-7695-3519-7/09 $25.00 © 2009 IEEE DOI 10.1109/CAR.2009.93
2
Combining (1) and (2), we have 2
2
Z1 − X 2 = Z 2 − B2 + 2BX − X 2
(4)
Therefore, the x-coordinate of the object P can be given by: 2
2
Z − Z2 + B2 X= 1 2B
(5)
Similarly, combining (2) and (3), we have 2
2
Z 3 = Z 2 − Y 2 + ( A − Y) 2
(6)
From (6), we can obtain the y-coordinate of the object P: 2
Y=
2
Z 2 − Z3 + A 2 2A
437
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
(7)
A. System Architecture Figure 2 shows the proposed ultrasonic-based localization scheme, where the target object is a mobile robot incorporating an ultrasonic sensor U0 as a transmitter. A micro-controller, H8tiny-3694 (Renesas Technology Corp.), is embedded into the robot for activating ultrasonic transmitter and communicating with its Zig-Bee module. Experiment environment also includes four poles forming a square around the measuring site, where an ultrasonic sensor is built on the top of each pole as a receiver. Based on ZigBee modules integrated with the ultrasonic sensors, a wireless network can be established for seamless communication among the ultrasonic sensors. As a result, the main controller PC can promptly give instructions to the ultrasonic sensors to transmit and receive signals. Distance measurements collected by 4 receivers of the measuring site are processed by the remote PC under an established 2D coordinate model to estimate the position of the robot.
U3
U4 Z3 P(X,Y)
U0 A Z1
Z2
Y
U1
U2
X
B
B. Ultrasonic Reflection Due to the angle limitation of ultrasonic signal, the detection capability will be affected, and it further limits the measurable area. In this paper, we try using a cone with 45° incline to reflect the incident ultrasonic signal from the ultrasonic transmitter to form a 360° plane, such that 4 ultrasonic sensors as receivers can easily receive the transmitting ultrasonic signal everywhere the mobile object moves in a detection area. Moreover, it is unnecessary for the ultrasonic transmitter to adjust its transmitting direction when one ultrasonic receiving sensor switches to the other one. Figure 3 illustrates how the transmitting ultrasonic signal forms a 360° ultrasonic plane via cone’s reflection.
Figure 1. A simplified localization scheme using ultrasonic-based method.
Comparing the measurement results obtained with predefined conditions, we can evaluate the feasibility of the coordinates if abnormal measurements occur. Thanks to these procedures for screening the measurements, accuracy of the coordinate of the object P can be significantly improved. III.
IMPROVED ULTRASONIC-BASED LOCALIZATION SYSTEM
To precisely estimate the position of a target object, for example, the mobile robot, distance measurements need to be obtained via the ultrasonic sensors. Subsequently, the distance measurements are used to calculate the coordinate of the robot based on an established two-dimensional (2D) coordinate model via a triangular relationship described as Section 2. In this system, in addition to ultrasonic sensors, Zig-Bee wireless modules are also required to transmit the related position information to the main controller PC. Furthermore, using ultrasonic reflection method can improve the localization performance due to solving the angel limitation of ultrasonic signal. U4
C. Operations At initial state, the main controller PC issues a command to 4 ultrasonic sensors at the receiving ends to prepare for receiving signals. At the same time, a command is issued to a specific ultrasonic sensor at the transmitting end on the robot via the Zig-Bee wireless network for transmitting ultrasonic signal. Once the ultrasonic sensors at the receiving ends receive signals, conversion into distance measurements is performed so that distance measurement can be conveyed to the main controller PC for further processing. Based on the collected distance measurements, position of the robot can then be calculated according to a 2D coordinate model.
U3
Ultrasonic sensor as transmitter
U0
Remote PC
U2
U1 : Ultrasonic signal
: Zig-Bee signal
: Ultrasonic sensor
: Ultrasonic sensor
as receiver
Cone for refraction : Incident wave : Reflection wave
as transmitter
Figure 3. Ultrasonic reflection for an ultrasonic transmitter.
Figure 2. Proposed ultrasonic-based localization scheme.
438
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
D. Zig-Bee Wireless Network The Zig-Bee module for each ultrasonic sensor, ZIG100B, provides three communication modes, including Peer to Peer, Waiting, and Broadcasting. In this paper, the Broadcasting mode is adopted because each receiver on the pole passively receives commands from the remote PC, which is suitable for broadcasting mode. Its simplicity for configuration and implementation is also an advantage.
Start
Output the localization results
Initialize (n=1)
Calculate position of the robot via a 2D coordinate m odel
nth ultrasonic sensor at the receiving end waits for receiving signals
IV.
Yes Polling com pleted? (n=4?)
Activate ultrasonic transmitter to transm it signals
Proceed with next ultrasonic sensor n=n+1
nth ultrasonic sensor at the receiving end receives and stores distance measurements
No
In our experiments, 9 points uniformly distributed in the measuring environment are sampled for testing. Experiment results based on the proposed ultrasonic-based localization are listed in Table 1. We find that the mean errors of the measurements lie within an acceptable range of ± 10cm, and the entire measurement area also can be enlarged to 300cm2 by reflecting the transmitting ultrasonic signal. Note that distance measurements are collected by way of polling in this paper so as to avoid interference that might have occurred. As a result, each receiver is activated to receive signal one at a time in a sequential order. Data collection efficiency is inevitably slowed down because of the polling process. For partially solving this problem, the independent remote PC in this proposed localization scheme is responsible for constantly consolidating distance measurements to calculate 2D coordinates of the robot.
Figure 4. Operations of the ultrasonic-based localization system. TABLE I.
Point Number
EXPERIMENT RESULTS OF PROPOSED LOCALIZATION SYSTEM
P(X,Y) (cm) Real Measured Coordinate Coordinate
Distance Error (cm)
1
(100,100)
(101,99)
(1,1)
2
(200,100)
(196,101)
(4,1)
3
(300,100)
(295,103)
(5,3)
4
(100,200)
(100,198)
(0,2)
5
(200,200)
(198,197)
(2,3)
6
(300,200)
(293,198)
(7,2)
7
(100,300)
(101,296)
(1,4)
8
(200,300)
(198,294)
(2,6)
9
(300,300)
(296,294)
(4,6)
EXPERIMENT RESULTS
V.
CONCLUSIONS
In this paper, an ultrasonic-based localization system for mobile robot is proposed to achieve a satisfactory accuracy in position, and solve the angle limitation of ultrasonic signal by using ultrasonic reflection. To provide seamless communication while avoiding interference in the measuring environment, Zig-Bee modules are integrated into all ultrasonic sensors for transmitting commands and receiving measurements by a sequential control. Based on an established 2D coordinate model, extensive experiments conducted have shown a satisfactory accuracy of ± 10cm in measuring the coordinates of the mobile robot can be obtained via the proposed localization scheme. When working with existing indoor localization methods, which are capable of identifying target objects in a larger area, the proposed localization scheme can further estimate the 2D coordinate of the object in better precision. Further extension of this work includes fusion of sensor data so as to provide more accurate localization results.
The ultrasonic sensors receive ultrasonic signal at the same frequency of 40k Hz. As a result, interference is inevitably encountered when all the 4 receivers receive signals at the same time. To avoid the interference, a sequential operation to control each ultrasonic sensor is employed in this work. Figure 4 shows the operation flow chart for this proposed ultrasonic-based localization system. To complete a polling cycle, ultrasonic sensors U1-U4 first enter a state of waiting to receive signal. Depending on commands received, a specific transmitter U0 is responsible for transmitting signal so that 4 receivers can detect and collect the measurements based on the time-of-flight (TOF). After a polling cycle is completed, 4 sets of distance measurements are ready for processing to calculate the coordinate of the object via a 2D coordinate model.
REFERENCES [1]
[2]
D. Tan, Q. Wang, Y. Wang, and R. Cao, “Research on position technology of indoors automobile robot,” Machinery & Electronics, vol. 9, pp. 46-47, 2005. J. Tang, T. Bao, and Z. Cai, “A localization method for mobile robot with optic mice in indoor environment,” Control & Automation, vol. 5, no. 21, pp. 20-22, 2005.
439
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
[3]
[4]
[5] [6]
[7] [8]
[9]
Y. H. Mu, Y. X. Yin, X. Y. Tu, and T. Q. Sun, “New real-time localization and motion planning method for indoor mobile robot,” Application Research of Computers, vol. 8, no. 24, pp. 106-108, 2007. K.E. Jianhua, H. Shen, and X. Wei, “Study on the ZigBee technology-based personnel position system of the underground coal mine,” Modern Electronics Technique, vol. 23, pp. 12-14, 2006. P. Enge and P. Misra, “Special issue on global positioning system,” Proceedings of the IEEE, vol. 87, no. 1, 1999. Y. Li, D. Xu, F. Li, M. Bai, and S. Zhang, “Application of GPS technology in agricultural land leveling survey,” Transactions of The Chinese Society of Agricultural Engineering, vol. 1, no. 21, pp. 6670, 2005. C. Shi, “Design of GPS vessel monitoring system,” Journal of Wuhan University of Technology, vol. 4, no. 26, pp. 448-450, 2002. Z. B. Feng, Y. S. Hua, X. Z. Qing, W. Y. Hua, and Z. Y. Zhong, “Application of GPS to the measurement of the roads,” Acta Metrologica Sinica, vol. 2, no. 23, pp. 98-105, 2002.
[10]
[11]
[12]
[13]
[14]
G. Ning, “GPS technique and its application in railway transportation engineering,” Journal of Changsha Railway University, vol. 3, no. 17, pp. 7-11, 1999. Y. Sun and P. Z. Fan, “RFID technology and its application in indoor positioning,” Computer Applications, vol. 5, no. 25, pp. 1205-1208, 2005. M. H. Zhang, S. S. Zhang and J. Cao, “Received-signal-strengthbased indoor location in wireless LANs,” Computer Science, vol. 24, no. 6, pp. 68-75, 2007. K. Zhang and G. Liu, “Research on a method of improving ultrasonic ranging precision,” Modern Electronics Technique, vol. 15, pp. 139-141, 2007. Y. L. Zhu, “Parameter research on ultrasonic orientation electrocircuit system,” Development & Innovation of Machinery & Electrical Products, vol. 19, no. 2, pp. 128-130, 2006. M. Piasecki, “Global localization for mobile robots by multiple hypothesis tracking,” Robotics and Autonomous Systems, vol. 16, no. 1, pp. 93-104, 1995.
440
Authorized licensed use limited to: St Johns University. Downloaded on April 7, 2009 at 20:45 from IEEE Xplore. Restrictions apply.
