AN EXPERT SYSTEM FOR GRINDING PROCESS OPTIMIZATION

The paper describes the development of an expert system based on a grinding database and a knowledge base, which include the results and conclusions of academic research as well as industrial expertise. For a given grinding situation, the system uses a model which takes into account a number of factors dependent on the input conditions. The value of these factors is determined by the inference engine. Thus, it relates the grinding conditions (e.g., wheel type, grinding machine, workpiece material, operating parameters) to the output parameters, such as grinding force and specific energy; it also provides data on more practical industrial parameters, such as surface roughness and probability of surface bums. The difference between the predicted and actual output parameters assessed in production can be used to improve the model. Examples of application of this system are given in the paper.