神经网络在采掘接替专家系统中的应用研究

Study on Application of Neural Network in Expert System of Mining Excavation Succession

为了实现采掘接替计划的自动生成与动态调整,构造了一种基于神经网络的采掘接替专家系统的结构模型。以矿井生产过程中存在的采掘失衡问题作为研究对象,提出了基于神经网络的知识获取方法,并对专家系统进行优化。该系统不但解决了专家系统获取知识的瓶颈问题,而且提高了整个系统的响应速度和准确性。采用ASP.NET和SQL Server等技术实现了专家系统的开发,并与人工编制的采掘计划进行比较。结果表明,基于神经网络的采掘接替专家系统能够更合理地生成采掘接替计划。

In order to realize automatic generation and dynamic adjustment for mining excavation succession plan, the structural model of mining succession expert system based on neural network is built. With the imbalances existing in mining production process as the research object, the knowledge acquisition method based on neural network is also proposed to optimize the expert system. The bottleneck of acquiring knowledge in expert system is solved and the response speed and precision of the whole system are improved. The expert system is developed by adopting various technologies, e. g. , ASP, NET and SQL server, etc. , and the comparison with manually compiled excavation plan is conducted. The results show that the proposed mining succession plan expert system based on neural network may more rationally generate excavation succession plan.