论文部分内容阅读
人工智能是工业过程控制和监控的极好工具。全世界已经报道了数以千计的工业应用实例。最近 ,人工智能系统的设计人员已经开始将智能技术、专家系统、模型逻辑和神经网络加以组合应用 ,以强化人工智能系统的功能。经证实专家系统是理想的候选系统 ,特别是用于矿物选别过程的控制。由于拥有不少成功的工程案例 ,本文介绍了一个用神经网络分析法对选厂给矿类型进行在线分类的专家系统。此专家系统的主要特点是 ,不同类型给矿可在选厂分类 ,并可采用不同的控制策略。除了分类以外 ,专家系统还设有一个能确定给矿类型的信息数据库。自学习数据库可扫描过程历史数据 ,并为正在处理的矿石类型推荐出最好的处理方法。该系统已经在芬兰奥托昆普公司的希土拉矿进行了试验。这种技术在其它工程项目的应用正在进行之中 ,这方面已有报道。
Artificial intelligence is an excellent tool for industrial process control and monitoring. Thousands of industrial applications have been reported worldwide. Recently, designers of artificial intelligence systems have begun to combine intelligent technologies, expert systems, model logic and neural networks to reinforce the capabilities of artificial intelligence systems. Proven expert systems are ideal candidate systems, especially for the control of mineral sorting processes. Owing to a number of successful engineering cases, this article presents an expert system for the online classification of ore dressing types by using neural network analysis. The main feature of this expert system is that different types of ore mines can be sorted at the mill and different control strategies can be adopted. In addition to the classification, the expert system also has a database of information that identifies the type of deposit. Self-learning databases scan process history data and recommend the best approach for the type of ore being processed. The system has been tested at Outokumpu’s Setila mine in Finland. The use of this technology in other engineering projects is underway, as has been reported in this area.