Ein Beitrag zur untertägigen Navigation von mobilen Maschinen; 1. Auflage
Hartmann, Tobias · RWTH Publications (RWTH Aachen) · 2021
The use of new technologies makes it possible to increase automation in primary raw material extraction and contributes to meeting growing demands for higher productivity, reducing environmental impact and improving safety. One of the main technical challenges is the (increased) automation of navigation of mobile machines in underground mining environment. Until now, automated navigation in underground mines has been possible to a very limited extent. Current developments in the field of sensors for localisation and environment detection, algorithms for environment mapping and recognition, motion planning and the performance of computer and network technology open up potential for increasing the automation of navigation. The developments are available in principle, but have only been tested in individual applications or laboratory environments for practical use and have not been integrated into overall systems in a practical manner. The navigation of a mobile machine can be broken down into different tasks. These tasks include localisation, mapping, route planning, environment recognition, trajectory planning and vehicle motion control. For the navigation tasks, the characteristics of the underground mining environment are presented, which illustrate the technical challenges of underground navigation. The increase in automation is always linked to a reduction in human action. For the objective classification of the automation of technical systems, different levels of automation are defined. Existing levels of automation and levels of driving automation do not, however, reflect the complexity of underground navigation or the possibilities of including auxiliary systems for navigation automation. In order to classify the automation of under-ground navigation, new levels of navigation automation degrees are defined for the mining industry and are referred to as Level of Mine Navigation Automation (LoMNA) in accordance with the driving automation standards. These range in six levels from no automation (LoMNA 0) to full automation (LoMNA 5). Existing systems from production and research do not exceed the automation degree of conditional automation (LoMNA 3). This is due to the insufficient recognition of the environment and non-existent or inflexible reaction to a changing (dynamic) environment. Based on the tasks of navigation and LoMNA, a comprehensive system concept is being developed for the first time in analogy to the human information processing model. This is developed into a system architecture based on available sensors and in principle suitable, novel algorithms and implemented in a novel navigation sys-tem. The navigation system is evaluated in the underground environment of an active mining operation, in certain scenarios based on the tasks of navigation. Apart from the first-time use of some of the sensors and algorithms in the underground environment, novel developments include, in particular, the infrastructure-free, map-based, geo-referenced localisation, the way in which the detection of and dynamic reaction to obstacles is implemented, and the testing of route and trajectory planning in the underground environment. The navigation system developed within this thesis demonstrated a novel level of autonomous, flexible, dynamic navigation in the underground mining environment. It reaches the level of high automation (LoMNA 4) and can thus be classified higher than all known systems. In future, the application-oriented further development of system automation is to be seen in the context of a broader range of requirements from environmental conditions to becoming an innovation. Furthermore, the investigation of the possibilities of integrating perceptive intelligent infrastructure for navigation automation and system interoperability for process integration are to be presented as approaches for further research and development.