Integrated Robot-Human Control in Mining Operations
This report contains a detailed description of the work conducted for the project on Integrated Robot-Human Control in Mining Operations at University of Nevada, Reno. This project combines human operator control with robotic control concepts to create a hybrid control architecture, in which the strengths of each control method are combined to increase machine efficiency and reduce operator fatigue. The kinematics reconfiguration type differential control of the excavator implemented with a variety of 'software machine kinematics' is the key feature of the project. This software re-configured excavator is more desirable to execute a given digging task. The human operator retains the master control of the main motion parameters, while the computer coordinates the repetitive movement patterns of the machine links. These repetitive movements may be selected from a pre-defined family of trajectories with different transformations. The operator can make adjustments to this pattern in real time, as needed, to accommodate rapidly-changing environmental conditions. A working prototype has been developed using a Bobcat 435 excavator. The machine is operational with or without the computer control system depending on whether the computer interface is on or off. In preparation for emulated mining tasks tests, typical, repetitive tool trajectories during surface mining operations were recorded at the Newmont Mining Corporation's 'Lone Tree' mine in Nevada. Analysis of these working trajectories has been completed. The motion patterns, when transformed into a family of curves, may serve as the basis for software-controlled machine kinematics transformation in the new human-robot control system. A Cartesian control example has been developed and tested both in simulation and on the experimental excavator. Open-loop control is robustly stable and free of short-term dynamic problems, but it allows for drifting away from the desired motion kinematics of the machine. A novel, closed-loop control adjustment provides a remedy, while retaining much of the advantages of the open-loop control based on kinematics transformation. While this technology has been tested extensively within the limitations of the project, the objective of full evaluation cannot be made. As only two people have operated the experimental excavator under transformed control in a digging condition, only a subjective evaluation of the technology can be made. Additionally, very limited data has been collected relating to energy benefits. Therefore, any statements that can be made about the potential for energy savings using this control architecture are widely based on a subjective evaluation of the performance. However, it is felt that an energy savings of around 20% can be expected using the new control technique effectively making the machine move easier and faster along a desired digging trajectory. In addition, the implementation of this technology greatly improves the ease of use of the excavator and reduces operator learning time and fatigue. Further work is also suggested to investigate problems encountered and associated with working envelope limitations and the very limited control resources provided in the experimental excavator designed and built originally for manual control only. These are discussed at the end of this report.
- Research Organization:
- Board Of Regents,Uccsn, Obo The University Of Nevada, Reno
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- FC26-04NT42087
- OSTI ID:
- 988569
- Country of Publication:
- United States
- Language:
- English
Similar Records
INTEGRATED ROBOT-HUMAN CONTROL IN MINING OPERATIONS
Human-machine cooperative telerobotics