Research On The Modeling Of Ceramic Grinding Force Based On The BP Neural Networks And Genetic Algorithms

Ceramic material has special features of hardness and brittle. As a vital parameter in grinding, grinding force is used to evaluate the grinding performance of an engineering ceramics, and the quality & damage degree of its surface. In order to exactly and effectively control the influence on grinding force during the grind process, it is very important to know every parameter related to grinding force. What is researched in this essay is how to build up the mathematics model of grinding force. Main conclusion and creative points read as followed:1. Utilize self-study characteristic of manual Neural Networks, build up the forecasting models of grinding force based on BP Neural Networks with 5 factors. In the model, besides considering the grinding conditions, the models considers the effect on grinding force based on engineering ceramic material performance and granularity of grinding wheel.2. By disposing and multidot searching of parameter muster based on genetic algorithm, only utilizing the advantages of evaluating gene unit based on adaptability function and specific searching direction, according to a lot of experimental analyse and comparison, restrict each parameters, build up the forecasting models of grinding force. By consulting lots of data, find few of building up the forecasting models of grinding force based on genetic algonrithm. So the forecasting models has innovation on some aspect.3. Do experiments on Si₃N₄ ceramic material, Al₂O₃ ceramic material, and ZrO₂ ceramic material. Provide samples for building up the forecasting models of grinding force based on BP Neural Networks and genetic algorithm, and make certain building up the forecasting models are doable.