Principle and Application of Blasting Expert System

Author: Guo Lianjun et al.
Publisher:
Publish Date: 1998-05-01
Features:
This book provides a detailed review of blasting optimization research and systematically introduces the basic principles and methods of blasting expert systems. It successfully integrates expert systems, artificial neural networks, and other related theories and methods for the development of blasting expert systems. It comprehensively analyzes and introduces knowledge representation, system reasoning, machine learning, and the implementation of blasting computer-aided design (CAD). Finally, it elaborates on the functions and usage methods of the blasting expert system software package with an application example. This book serves as a reference for science and technology personnel engaged in mine blasting design and production, and can also be used as a textbook or reference book for senior undergraduate or graduate students in mining, blasting, transportation, geotechnical engineering, and related fields.
Excerpt:
With the development and application of computer technology and other high technologies, mine blasting research has entered a new phase. Due to the complexity of mine blasting and the high cost of practical experiments, computer simulation methods and computer-aided design have emerged rapidly, bringing vitality to mining enterprises. Programs such as BlastingPlanDesigner, Blaspa, and BlastCAD are representative examples of computer simulation and computer-aided design. This work continues to advance today. In addition to simulation methods and CAD applications, artificial intelligence theory and technology are also being applied to mining, and plans aimed at intelligent mine technology are underway. An intelligent mine is one that applies computer technology to achieve information and data collection, bidirectional communication, and real-time control across the entire mine. Blasting, a process involving a large amount of uncertain information, undoubtedly provides a broad field for the application of artificial intelligence systems.
1.3.2 Differences Between Blasting Expert Systems and Traditional Computer-Aided Design
The basic structure of most traditional computer-aided design systems is the same: data + algorithms = programs. Based on the selected computational model, corresponding computer programs are developed. These models may be physical models or empirical models of blasting, both of which utilize the high computational power of computers to seek blasting design solutions that meet quality requirements. The computational process here is flawless, and the results are precise. As long as the required data is input, the computer will always produce corresponding results, regardless of whether the data aligns with the actual results. Unfortunately, these models cannot fully encompass all aspects of blasting work, and more notably, these systems require exact information. Once certain uncertainties are introduced into the program, it can lead to incorrect computational results. In open-pit mining, uncertainties and ambiguous information are abundant and unavoidable. Therefore, in traditional methods, either certain uncertainties are ignored, or assumptions are used as substitutes, resulting in the simplification of ambiguous factors into mere quantities and the simplification of complex processes into models, all of which introduce incalculable errors.
Expert systems, however, can, like human experts, draw the best conclusions based on incomplete or uncertain information, using accumulated experience and knowledge through analysis and inference. Additionally, through system operation, they can continuously enrich and refine their knowledge, enhancing their functionality as they process more blasting problems. In other words, traditional methods rely on precise calculations to obtain results, while expert systems rely on rigorous reasoning processes to reach conclusions. Furthermore, expert systems, like humans, can learn from experience, accumulate data, and correct outputs, which is particularly important for blasting. Due to the complexity of geological conditions, in open-pit blasting, various special conditions are frequently encountered, such as faults and fractured zones. The handling of these special situations relies on the experience of experts, which is often difficult for ordinary computer programs to achieve. If an expert system establishes a special knowledge base and processing rules for handling various possible special situations, it can accomplish this task, and with the accumulation of knowledge, its ability to handle special situations will continue to strengthen. Additionally, the knowledge base of an expert system differs from traditional blasting databases. Traditional blasting databases primarily serve management functions, including queries, retrievals, statistics, and printing. In contrast, the database in an expert system can be a database, a description of objective laws expressed in certain rules, or a weighted matrix of network connections. Knowledge can be both theoretical and empirical, with experience being refined and analyzed by most experts and not randomly accumulated. Once knowledge is loaded into the knowledge base or integrated into the network, it can serve as a basis for the next blasting design. Moreover, the knowledge in the knowledge base, like human knowledge, can be continuously updated, modified, and supplemented to improve the knowledge level.
1.3.3 Research and Progress of Blasting Expert Systems at Home and Abroad
Expert systems and neural networks are relatively new research fields, and their application in mining, especially in mine blasting, is still limited. Some countries have applied expert systems in mine exploitation and tunneling, achieving significant progress. For example, over 50 mines in the United States, the United Kingdom, and Canada have applied expert systems, which have been continuously developed and tested over the years to achieve perfection. The adoption of expert systems has significantly reduced the cost of blasting and drilling in mines.