The main objective of the Intelligent Network-Based Control Project is to develope a wireless unmanned vehicle to find its way through a platform with obstacles. It should arrive at the destination specified by a user in a remote location using the shortest amount of time, and avoiding any collisions.

Main focuses are:

• How to recognize the vehicle and obstacles.
• How to generate a path around the obstacles in the shortest amount of time.
• How network delays contribute to control errors.

Fig. 1 Intelligent Space via Network-Based Control.

The vision of the vehicle is established by a wireless network camera placed on top of the platform. The vehicle is battery operated and completely controlled by wireless communication via IP network. Image processing technology, as well as hardware and software implementations are utilized to demonstrate the path generation and path tracking algorithms. Gain Scheduler Control algorithm may also be implemented in this project to compensate for any network delay disturbance.

Fig. 2 Path generation algorithm.

Journal Papers

• M.-Y. Chow, Y. Tipsuwan, “Gain Adaptation of Networked Dc Motor Controllers on QoS Variations,” IEEE Transactions on Industrial Electronics, Vol. 50, no. 5, October, 2003.
• Y. Tipsuwan and M.-Y. Chow, "Control Methodologies in Networked Control Systems," Control Engineering Practice, vol. 11, 2003, pp.1099-1111.

Conference Papers

• Y. Tipsuwan, M.-Y Chow, “Neural Network Middleware for Model Predictive Path Tracking of Networked Mobile Robot over IP Network,” IEEE IECon’03, Roanoke, VA, Nov 2 – Nov 6, 2003.
• Y. Tipsuwan, M.-Y. Chow, “An Implementation of a Networked PI Controller over IP Network,” IEEE IECon’03, Roanoke, VA, Nov 2 – Nov 6, 2003.
• Y. Tipsuwan, M.-Y. Chow, “On the Gain Scheduling for Networked PI Controller Over IP Network,” 2003 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, Port Island, Kobe, Japan, July 20-24, 2003.
• Y. Tipsuwan, M.-Y. Chow, “Gain adaptation of mobile robot for compensating QoS deterioration”, Proceedings of IECon’02, Sevilla, Spain, November 5 – 8, 2002.
• Y. Tipsuwan, M.-Y. Chow, “Network-Based Controller Adaptation Based On QoS Negotiation and Deterioration,” IECon01, Denver, CO, Nov.28-Dec.02, 2001, pp. 1794 -1799.
• M.-Y. Chow, Y. Tipsuwan, “Network-based control adaptation for network QoS variation,” MILCOM 2001, October 28-31, 2001, McLean, VA, pp. 257-261.

Thesis and Technical Reports

• B. Ayhan, W.L. Leung, Z. Li, T. Richards, R. Vanijjirattikhan, L. Xu, "iState--Optimal Control Project Report", ECE 726, Department of Electrical and Computer Engineering, North Carolina State University, May 2004.

Tutorials

• M.-Y. Chow, “Methodologies in Time Sensitive Network-Based Control Systems,” 2003 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, Port Island, Kobe, Japan, July 20-24, 2003.
• M.-Y. Chow, Y. Tipsuwan, “Real Time Network-Based Control System,” IEEE IECon 2002 Tutorial, Sevilla, Spain, November 5, 2002.
• M.-Y. Chow, Y. Tipsuwan, “Network-Based Control Systems: A Tutorial,” Proceedings of IEEE IECon 2001 Tutorial, November 28 – December 2, Denver, CO, pp. 1593 -1602.

Software prototype:

In this project, we followed the fast prototyping concept in order to deliver the first complete prototype as soon as we can. The project members are allowed to develop the program in their parts with their preference in choosing the programming language between C and Java. The overall parts will be combined together by using CORBA technology. As a result, we extensively use different kinds of the development tools for rapid implementation. Most of the development tools we chose are free and can be downloaded from the Internet.

• Image Acquisition (for storing real-time picture into image files)
• Image Processing (for the position acquisition of the UV and the obstructions)
• Path Generation (for generating an optimal path)
• Gain Scheduler Middleware (for providing control signals to drive the UV according to the communication network environment; this module only has the UV path tracking for now)

Hardware prototype:

The unmanned vehicle used in this project, Johnny6, is composed of various pieces of hardware, including a PC104 embedded computer, a hard drive, a wireless card and adapter, a custom built interface card, and a chassis with voltage regulators, batteries, and DC motors.

The total weight of Johnny6 is 15lb., most of which are contributed by the three batteries, which power up the vehicle and the on-board computer for over an hour. Control signals are recieved via IP using a 802.11b wireless card. The on board computer then calculate these control signals and turn them into pause-width-modulation signals to control each DC motor. After each 700ms, the main controler in the server will recieve the new position of the vehicle and calculate a new control signal to be sent via wireless channel.