论文部分内容阅读
在齿轮坯多工位锻造生产中,齿轮坯预锻件的设计直接影响到终锻时金属流动、锻模型腔的充满情况、锻件质量和模具寿命。本文利用BP神经网络处理高度非线性问题的特点,建立了预锻件尺寸和终锻成形力、终锻最大模具应力之间的网络模型,再结合遗传算法的全局寻优功能,确定了在终锻成形力最小以及终锻最大模具应力最小时的最佳的预锻件形状和尺寸。
In the multi-station gear blank forging production, the design of the gear preform forging directly affects the flow of metal during the final forging, the filling cavity of the forging die, the quality of the forging and the life of the die. In this paper, by using BP neural network to deal with the characteristics of highly nonlinear problems, a network model of pre-forging size, final forging forming force and maximum final die forging stress was established. Combined with the global optimization function of genetic algorithm, The best shape and size of the preform when the forming force is minimum and the final die forging the maximum die stress is the minimum.