针对半固态挤压复合材料棒材成形时工艺参数难于选取、试验工作量大的问题 ,采用人工神经网络技术与试验相结合的方法 ,通过对样本处理、神经网络模型参数及收敛性等进行分析 ,建立了工艺参数ANN预测模型 ,可以对复合材料半固态挤压成形的关键工艺参数进行预测 ,预测值与试验值吻合较好 ,最大误差不超过 0 72 % ,为复合材料半固态挤压成形的应用开辟了有效途径 Aiming at the difficulty in selection of process parameters and large amount of test work during the forming process of semi-solid extruded composite bar, the method of combining artificial neural network and experiment is adopted to analyze the parameters of the sample processing, neural network model and convergence , The ANN prediction model of process parameters was established to predict the key process parameters of semi-solid extrusion of composite materials. The predicted values ​​agree well with the experimental values, the maximum error is no more than 0 72% The application has opened up an effective way