Research On NC Cutting Parameter Optimum Matching Expert System

One of the characteristics of modern manufacturing industry is the multiple-varieties and small-batches production comparing previous single-variety and large-batches one, which brings everlastingly higher demand on the flexibility of the manufacturing system. For enterprises, they have to face the challenges of larger investment in equipments and higher costs. Among all the solutions, machining parameter optimization is a valid one. By integrating various computer technologies such as group evolution optimization theory, AI and database, this paper constructs a multi-objective metal cutting parameter optimization expert system, which explores a new method for metal cutting parameter optimization. The content of the paper includes the following:1. Introduces the fuzzy theory to the traditional metal cutting parameter optimization model. Within wholly consideration of all the possible uncertainties of the metal cutting process, exampled by milling process, a multi-objective cutting parameter fuzzy optimization model is established.2. Introduces Simulating Annealing Genetic Algorithm as the optimizing engine, whose advantages are illustrated by a calculating case, and realize the optimization of the previous fuzzy model.3. Setup the metal knowledge acquisition model of metal cutting parameters based on Artificial Neural Network (ANN). By studying the optimized parameters set the knowledge acquisition model obtains cutting parameter matching knowledge under corresponding circumstance, so that when the altering of the cutting condition occurs, the model can work out a satisfactory optimized parameter without interference of the fuzzy model, whose operation needs large computation cost. This knowledge acquisition model also establishes a foundation for the construction of the knowledge matching database of the expert system.4. Develop the prototype software of the expert system, which takes the multi-objective metal cutting parameter fuzzy model as knowledge source, utilizes a combined strategy between rule-based knowledge database and ANN database as knowledge expression method. A friendly man-machine interface is provided, and multi- and single-objective metal cutting parameter optimization is realized. The system provides proper and matching parameters of cutters and cutting conditions, so that the expert knowledge of cutting field can be fully used to instruct real production,to reach the goal of raising the utilization efficiency, shortening of manufacturing cycle and elevation of productivity.