Pittsburgh, PA: U.S. Department of the Interior, Bureau of Mines, TN 394, 1992 Feb; :1-2
Objective: Track and control the tram maneuvers of continuous mining machines by developing and evaluating a laser sensor system. Background: Progressive technology has the potential to greatly improve mine-worker safety, helping to create a safer, better working environment for mine workers by keeping them away from the dangerous face area. To benefit from such technology, the Bureau is investigating remote, computer-assisted operation of mining systems for underground, highwall, and surface mining applications. A cornerstone for computer-assisted operation is an effective guidance system. Since there are no guidance systems on any commercial continuous miners (CM's), the Bureau itself developed and tested a laser-based sensor system, based on a commercial sensor called Lasernet. The laser-based guidance system employs four laser scanning sensors at fixed locations in a mine entryway to report the angular coordinates of two cylindrical retroreflective targets mounted on a CM. Using triangulation, a real-time microcomputer system processes the sensor data and updates the xy position and heading (h) of the CM five times per second. Using a simple control algorithm and communications interface to the CM control computer, the system also controls rotational and translational tram maneuvers. Approach: Researchers first took the laser system underground in early 1991 and performed several experiments. The Lasernet sensors, mounted on tripods, are located near the tail end on both sides of the CM. Two cylindrical retroreflective targets are also mounted on the tail end of the CM. During underground experimentation, Lasernet sensors were housed in explosion-proof enclosures and mounted on two poles at fixed positions on both sides of the entryway. The most crucial factor in maintaining the laser system's reliability is keeping the targets within the Lasernet sensor's 110° field of view. While 110° easily covers an entryway, Lasernet sensor's target detection range has a 10.67 m limit. However, extended-range Lasernet sensors with a 30.48 m target detection range are currently available. Another problem occurs with an unusually irregular floor. Position and heading updates of the CM become unavailable when the scanning beam of the laser doesn't cross the targets, scanning cither above or below the target surface. As a solution, researchers used taller targets, increasing the vertical tolerance from .35 m to .46 m. They also added two Lasernet sensors, each mounted .15-.30 m higher than the original two sensors, thus increasing the system's vertical tolerance from .35 m to .76 m. An alternative solution would be to add a mechanical mirrored assembly to the sensor which would sweep the beam vertically when no targets are detected. Test Results: The data taken to determine the accuracy of the laser system in tracking position and heading and controlling the tram maneuvers of the Joy 14CM in both the Bureau's Mine Equipment Test Facility (METF) and underground were promising. The first experiment showed good accuracy in tracking the position and heading of the CM as it performed a variety of maneuvers underground. After each move, a transit reading verified the xy position and h of the CM (table 1). Average errors were: h, 0.4°; x position, 1.7 cm; and y position, .51 cm. The second experiment shows the accuracy in tracking and controlling the translations and rotations of the CM. Table 2 depicts the test results. Column one shows the desired tram maneuver. Column two shows the actual translation or rotation of the CM, calculated by using two transit measurements taken before and after the maneuver. Column three shows the translation or rotation measured by the laser system. The final columns show errors for determining Lasernet accuracy. Average measurement errors were 1.6 cm translation, 0.3° rotation. Average control errors were 5.2 cm translation, 0.7° rotation. Related underground experiments on computer-assisted mining systems used the laser system to track and control the tramming of the CM. The laser system was an integral part of two experiments to test the first fully-controlled, automatic execution of multiple sump-shear cutting cycles. The first experiment used a preprogrammed coal removal script created and executed by the autonomous mining research and development system (AMREDS), a window-based computer program developed by the Bureau and Carnegie Mellon University. The second experiment used Carnegie Mellon University's MINENA V software to dynamically plan the maneuvers and appendage commands necessary during a two-pass mining sequence. Details are omitted here for brevity. Results of the underground experiments show promise. The system tracks the position of the CM to an accuracy of better than 2.5 cm and the heading to better than 0.5°. Tram control accuracies are better than 10.2 cm translation, and better than 2° rotation. Additionally, the laser system can be applied to other mining equipment such as shuttle cars, roof bolters. Simple alterations would make the system capable of tracking a roof bolter maneuvering through the face area. Also, sensor and target locations could be switched. For example, by placing Lasernet sensors on-board the vehicle, and the targets along the path from the face to the dump point, it would be possible to guide shuttle cars.
Pittsburgh, PA: U.S. Department of the Interior, Bureau of Mines, TN 394