论文部分内容阅读
为了有效地控制激光铣削层质量,建立了激光铣削层质量(铣削层宽度、铣削层深度)与铣削层参数(激光功率、扫描速度和离焦量)的BP神经网络预测模型。采用粒子群算法优化了BP神经网络的权值和阈值,构建了基于粒子群神经网络的质量预测模型。所提出的PSO-BP算法解决了一般BP算法迭代速度慢,且易出现局部最优的问题,并以Al2O3陶瓷激光铣削质量预测为例,进行算法实现。仿真结果表明:提出的PSO-BP算法迭代次数大大减少,且预测误差明显减少。所构建的质量预测模型具有较高的预测精度和实用价值。
In order to control the quality of laser milling layer, a BP neural network prediction model of laser milling layer quality (milling layer width, milling depth) and milling layer parameters (laser power, scanning speed and defocus amount) is established. Particle swarm optimization algorithm is used to optimize the weight and threshold of BP neural network, and a quality prediction model based on PSO neural network is constructed. The proposed PSO-BP algorithm solves the problem of slow iteration speed and easy local optimization of the general BP algorithm. The algorithm is implemented by taking the quality prediction of Al2O3 ceramic laser milling as an example. The simulation results show that the proposed PSO-BP algorithm greatly reduces the number of iterations and significantly reduces the prediction error. The constructed quality prediction model has high prediction accuracy and practical value.