黏结漏钢作为最常见的漏钢事故,不仅会损坏连铸设备,甚至威胁操作人员生命安全。常规的漏钢预报模型主要依赖工艺参数的阈值进行判断和简单的统计分析,没有充分利用数据的时序变化,限制了模型的准确性。为了解决上述问题,将鲸鱼优化算法...黏结漏钢作为最常见的漏钢事故,不仅会损坏连铸设备,甚至威胁操作人员生命安全。常规的漏钢预报模型主要依赖工艺参数的阈值进行判断和简单的统计分析,没有充分利用数据的时序变化,限制了模型的准确性。为了解决上述问题,将鲸鱼优化算法(whale optimization algorithm,WOA)和长短期记忆神经网络(long short term memory,LSTM)结合,构建了一种基于深度学习的WOA-LSTM漏钢预报模型。提取温度特征、静态几何特征及动态特征,使用皮尔逊相关系数筛选出与漏钢事故相关性较高的特征参数,包括温升异常和温降区域温度变化率均值、最大值等11个特征。利用鲸鱼优化算法对长短期记忆神经网络的超参数进行寻优,以均方误差作为模型损失函数,通过循环迭代搜索出最优的网络超参数。在模型训练过程中,采用滑动窗口技术输入训练样本,使模型能够更好地学习和捕捉连铸过程中工艺参数的时序变化特征。最后使用某钢厂的实际生产数据进行了试验,与BP(back propagation)、LSTM及WOA-BP模型相比,WOA-LSTM预测模型在多个性能指标上均表现出色,能更精准地捕捉到特征数据的时序变化趋势,且模型的收敛速度快、预测精度高。该模型的报出率为98.4%,预报率为96.8%,能够满足钢厂实际生产的要求。展开更多
In order to solve the springback problem in sheet metal forming, the trial and error method is a widely used method in the factory, which is time-consuming and costly for its non-direction and non-quantitative. Finite...In order to solve the springback problem in sheet metal forming, the trial and error method is a widely used method in the factory, which is time-consuming and costly for its non-direction and non-quantitative. Finite element simulation is an e ective method to predict the springback of complex shape parts, but its precision is sensitive to the simulation model, particularly material model and boundary conditions. In this paper, the simple iterative method is introduced to establish the iterative compensation algorithm, and the convergence criterion of iterative parameters is put forward. In addition, the new algorithm is applied to the V-free bending and stretch-bending processes, and the convergence of curvature and bending angle is proved theoretically and verified experimentally. At the same time,the iterative compensation experiments for plane bending show that, the new method can predict the next compensaantido tnh ev atlaureg ebta cseurdv oatnu trhe ew sitphri tnhgeb earcrko ro fo fe laecshs ttehsat,n s0 o. 5 th%a ta rteh eo btatraigneet db aefntedri n2 g-3 a nitgelrea tiwoitnhs.t Thhei se rrreosre aorf clhe sps rtohpaons e±s 0 a.1%new iterative compensation algorithm to predict springback in sheet metal forming process, where each compensation value depends only on the iteration parameter di erence before and after springback for the same forming process of same material.展开更多
Sparse arrays of telescopes have a limited (u, v)-plane coverage. In this paper, an optimization method for designing planar arrays of an aperture synthesis telescope is proposed that is based on distributed genetic a...Sparse arrays of telescopes have a limited (u, v)-plane coverage. In this paper, an optimization method for designing planar arrays of an aperture synthesis telescope is proposed that is based on distributed genetic algorithm. This distributed genetic algorithm is implemented on a network of workstations using community communication model. Such an aperture synthesis system performs with imperfection of (u, v) components caused by deviations and(or) some missing baselines. With the maximum (u, v)-plane coverage of this rotation-optimized array, the image of the source reconstructed by inverse Fourier transform is satisfactory.展开更多
文摘黏结漏钢作为最常见的漏钢事故,不仅会损坏连铸设备,甚至威胁操作人员生命安全。常规的漏钢预报模型主要依赖工艺参数的阈值进行判断和简单的统计分析,没有充分利用数据的时序变化,限制了模型的准确性。为了解决上述问题,将鲸鱼优化算法(whale optimization algorithm,WOA)和长短期记忆神经网络(long short term memory,LSTM)结合,构建了一种基于深度学习的WOA-LSTM漏钢预报模型。提取温度特征、静态几何特征及动态特征,使用皮尔逊相关系数筛选出与漏钢事故相关性较高的特征参数,包括温升异常和温降区域温度变化率均值、最大值等11个特征。利用鲸鱼优化算法对长短期记忆神经网络的超参数进行寻优,以均方误差作为模型损失函数,通过循环迭代搜索出最优的网络超参数。在模型训练过程中,采用滑动窗口技术输入训练样本,使模型能够更好地学习和捕捉连铸过程中工艺参数的时序变化特征。最后使用某钢厂的实际生产数据进行了试验,与BP(back propagation)、LSTM及WOA-BP模型相比,WOA-LSTM预测模型在多个性能指标上均表现出色,能更精准地捕捉到特征数据的时序变化趋势,且模型的收敛速度快、预测精度高。该模型的报出率为98.4%,预报率为96.8%,能够满足钢厂实际生产的要求。
基金Supported by Hebei Provincial Natural Science Foundation of in China(Grant Nos.E2015203244,E2016203266)Program for the Youth Top-notch Talents of Hebei Province
文摘In order to solve the springback problem in sheet metal forming, the trial and error method is a widely used method in the factory, which is time-consuming and costly for its non-direction and non-quantitative. Finite element simulation is an e ective method to predict the springback of complex shape parts, but its precision is sensitive to the simulation model, particularly material model and boundary conditions. In this paper, the simple iterative method is introduced to establish the iterative compensation algorithm, and the convergence criterion of iterative parameters is put forward. In addition, the new algorithm is applied to the V-free bending and stretch-bending processes, and the convergence of curvature and bending angle is proved theoretically and verified experimentally. At the same time,the iterative compensation experiments for plane bending show that, the new method can predict the next compensaantido tnh ev atlaureg ebta cseurdv oatnu trhe ew sitphri tnhgeb earcrko ro fo fe laecshs ttehsat,n s0 o. 5 th%a ta rteh eo btatraigneet db aefntedri n2 g-3 a nitgelrea tiwoitnhs.t Thhei se rrreosre aorf clhe sps rtohpaons e±s 0 a.1%new iterative compensation algorithm to predict springback in sheet metal forming process, where each compensation value depends only on the iteration parameter di erence before and after springback for the same forming process of same material.
基金This project was supported by the High Technology Research and Development Programme of China (2002AA111040).
文摘Sparse arrays of telescopes have a limited (u, v)-plane coverage. In this paper, an optimization method for designing planar arrays of an aperture synthesis telescope is proposed that is based on distributed genetic algorithm. This distributed genetic algorithm is implemented on a network of workstations using community communication model. Such an aperture synthesis system performs with imperfection of (u, v) components caused by deviations and(or) some missing baselines. With the maximum (u, v)-plane coverage of this rotation-optimized array, the image of the source reconstructed by inverse Fourier transform is satisfactory.
