功率模块铜线键合工艺参数优化设计

Optimization design on process parameters of copper wire bonding for power modules

为了提高功率模块铜线键合性能,采用6因素5水平的正交试验方法,结合BP(Back Propagation)神经网络与遗传算法,提出了一种铜线键合工艺参数优化设计方案。首先,对选定样品进行正交试验并将结果进行极差分析,得到工艺参数对键合质量的影响权重排序。其次,运用BP神经网络构建了铜线键合性能预测模型,并通过遗传算法对BP神经网络适应度函数求解,得到了工艺参数的最优值。将BP-遗传算法与传统优化方法的优化结果进行对比,发现经BP-遗传算法优化后的铜线键合工艺稳定性提升更加明显。最后,对功率模块进行了功率循环试验,结果表明经BP-遗传算法优化后的模块功率循环能力得到显著提升。

In order to improve the copper wire bonding performance of the power modules,an optimization design scheme of process parameters for copper wire bonding was proposed by using the six-factor five-level orthogonal test method,and combining the back propagation(BP)neural network and genetic algorithm(GA).Firstly,the selected samples were orthogonally tested and the results were analyzed by range analysis to generate the influence weight ranking of the process parameters on the bonding quality.Secondly,a prediction model of copper wire bonding performance was constructed using the BP neural network,and the optimal values of process parameters were generated by solving the BP neural network fitness function with GA.Comparing the optimization results from the BP genetic algorithm with those from the traditional methods,the former was found in the stability of the copper wire bonding process improved more significantly.Finally,the power cycle test was carried out on the power module,showing that the power cycle capability of the module optimized by the BP genetic algorithm was significantly improved.