Scheduling Semiconductor Testing Facility by Using Cuckoo Search Algorithm With Reinforcement Learning and Surrogate Modeling
Scheduling Semiconductor Testing Facility by Using Cuckoo Search Algorithm With Reinforcement Learning and Surrogate Modeling
复制标题
DOI：
10.1109/tase.2018.2862380
复制
发表时间：
2019-04
影响因子：
5.6
通讯作者：
Zhengcai Cao;Chengran Lin;Mengchu Zhou;Ran Huang
中科院分区：
文献类型：
--
作者：
Zhengcai Cao;Chengran Lin;Mengchu Zhou;Ran Huang
本文考虑了一个具有多资源约束的半导体最终测试调度问题，并被证明是NP-hard问题。为了最小化此调度问题的完工时间，提出了一种具有强化学习 (RL) 和代理建模的布谷鸟搜索算法。在强化学习的基础上，提出了一种参数控制方案，以确保种群所需的多样化和集约化，根据Rechenberg的1/5准则，使用有益突变的比例作为反馈信息。为了降低计算复杂度，采用代理模型来评估解决方案的相对排名。提出了一种基于编码值相对排序和模函数的启发式方法，将 Lévy 飞行获得的连续解转换为离散解。给出了计算复杂度和收敛性分析结果。所提出的算法通过基准和随机生成的案例进行了验证。通过各种仿真实验以及所提出的算法与几种流行方法的比较来验证其有效性。从业人员注意——半导体最终测试的安排通常通过智能优化算法来解决。然而，它们大多数都是参数敏感的，因此，选择合适的参数是一个巨大的挑战。为了解决参数选择问题，我们提出了一种强化学习（RL）算法来自我调整其参数。为了减少计算负担，我们建议在强化学习中使用奖励函数的代理建模，并确定在布谷鸟搜索中应保留哪些巢。因此，我们的算法比现有的半导体最终测试设备算法具有更高的鲁棒性，并且可以获得高质量的时间表。此外，通过所提出的代理模型具有较低的计算复杂度，因此可以在短时间内获得可行的解决方案以进行实时调度。实验结果表明，该方法明显优于一些现有算法。因此，它可以很容易地应用于工业半导体最终测试设施调度问题。
A semiconductor final testing scheduling problem with multiresource constraints is considered in this paper, which is proved to be NP-hard. To minimize the makespan for this scheduling problem, a cuckoo search algorithm with reinforcement learning (RL) and surrogate modeling is presented. A parameter control scheme is proposed to ensure the desired diversification and intensification of population on the basis of RL, which uses the proportion of beneficial mutation as feedback information according to Rechenberg’s 1/5 criterion. To reduce computational complexity, a surrogate model is employed to evaluate the relative ranking of solutions. A heuristic approach based on the relative ranking of encoding value and a modular function is proposed to convert continuous solutions obtained from Lévy flight into discrete ones. The computational complexity and convergence analysis results are presented. The proposed algorithm is validated with benchmark and randomly generated cases. Various simulation experiments and comparison between the proposed algorithm and several popular methods are performed to validate its effectiveness. Note to Practitioners—Scheduling of semiconductor final testing is usually solved via intelligent optimization algorithms. Nevertheless, most of them are parameter-sensitive, and thus, selecting their proper parameters is a huge challenge. In order to deal with the parameter selection issue, we propose a reinforcement learning (RL) algorithm to self-adjust their parameters. To reduce the computational burden, we propose to use surrogate modeling of the reward function in RL and determine which nests should be reserved in cuckoo search. As a result, our algorithm possesses higher robustness and can obtain a high-quality schedule than the existing algorithms for semiconductor final testing facility. In addition, it has a lower computational complexity via the proposed surrogate model, and thus, a feasible solution can be obtained in a short time for real-time scheduling. Experimental results show that the proposed method well outperforms some existing algorithms. Hence, it can be readily applied to industrial semiconductor final testing facility scheduling problems.