为了确定油井或油井泵电机的性能下降，并为特定油井制定最佳维护计划，工程师必须评估几个月的油井产量和运行数据。使用一个专家系统，可以自动完成评估数月油井数据的重复而繁琐的过程。最终，应设计一个专家系统，自动处理月度油井数据，识别性能下降的油井特征，并建议具体的维护措施以提高性能。 本文介绍了专家系统开发的初始步骤，包括讨论问题对专家系统的适用性、专家系统中的知识组织、专家系统工具的选择、迄今为止的实际开发工作，以及此时可以得出的结论。

To identify degrading performance of wells or well pump motors, and to set up an optimum maintenance schedule for a particular well, an engineer must evaluate several months of data on well output and operation. The repetitive and tedious procedure of evaluating many months of well data can be automated using an expert system. Ultimately, an expert system should be designed to automatically process monthly well data, identify wells with characteristics that indicate degrading performance, and recommend specific maintenance actions to improve performance. This paper presents the initial steps in the development of the expert system, including discussions of the suitability of the problem for an expert system, knowledge organization in the expert system, the choice of an expert system tool for this task, actual development efforts thus far, and conclusions that can be drawn at this point.