The main aim of this research is to get a better knowledge and understanding of the micro-scale oscillatory networks behavior in the solid propellants reactionary zones. Fundamental understanding of the micro-and nano...The main aim of this research is to get a better knowledge and understanding of the micro-scale oscillatory networks behavior in the solid propellants reactionary zones. Fundamental understanding of the micro-and nano-scale combustion mechanisms is essential to the development and further improvement of the next-generation technologies for extreme control of the solid propellant thrust. Both experiments and theory confirm that the micro-and nano-scale oscillatory networks excitation in the solid propellants reactionary zones is a rather universal phenomenon. In accordance with our concept,the micro-and nano-scale structures form both the fractal and self-organized wave patterns in the solid propellants reactionary zones. Control by the shape, the sizes and spacial orientation of the wave patterns allows manipulate by the energy exchange and release in the reactionary zones. A novel strategy for enhanced extreme thrust control in solid propulsion systems are based on manipulation by selforganization of the micro-and nano-scale oscillatory networks and self-organized patterns formation in the reactionary zones with use of the system of acoustic waves and electro-magnetic fields, generated by special kind of ring-shaped electric discharges along with resonance laser radiation. Application of special kind of the ring-shaped electric discharges demands the minimum expenses of energy and opens prospects for almost inertia-free control by combustion processes. Nano-sized additives will enhance self-organizing and self-synchronization of the micro-and nano-scale oscillatory networks on the nanometer scale. Suggested novel strategy opens the door for completely new ways for enhanced extreme thrust control of the solid propulsion systems.展开更多
基金supported by the Western-Caucasus Research Center
文摘The main aim of this research is to get a better knowledge and understanding of the micro-scale oscillatory networks behavior in the solid propellants reactionary zones. Fundamental understanding of the micro-and nano-scale combustion mechanisms is essential to the development and further improvement of the next-generation technologies for extreme control of the solid propellant thrust. Both experiments and theory confirm that the micro-and nano-scale oscillatory networks excitation in the solid propellants reactionary zones is a rather universal phenomenon. In accordance with our concept,the micro-and nano-scale structures form both the fractal and self-organized wave patterns in the solid propellants reactionary zones. Control by the shape, the sizes and spacial orientation of the wave patterns allows manipulate by the energy exchange and release in the reactionary zones. A novel strategy for enhanced extreme thrust control in solid propulsion systems are based on manipulation by selforganization of the micro-and nano-scale oscillatory networks and self-organized patterns formation in the reactionary zones with use of the system of acoustic waves and electro-magnetic fields, generated by special kind of ring-shaped electric discharges along with resonance laser radiation. Application of special kind of the ring-shaped electric discharges demands the minimum expenses of energy and opens prospects for almost inertia-free control by combustion processes. Nano-sized additives will enhance self-organizing and self-synchronization of the micro-and nano-scale oscillatory networks on the nanometer scale. Suggested novel strategy opens the door for completely new ways for enhanced extreme thrust control of the solid propulsion systems.
文摘随着算力网络中计算资源与虚拟化设备的广泛应用,在算力网络虚拟化中,针对云集群弹性伸缩策略基于阈值的响应式触发过程中存在的弹性滞后问题,提出一种基于Transformer的预测式云集群资源弹性伸缩方法(Predictive Cloud Cluster Resource Elastic Scaling Method Based on Transformer,Cloudformer).该方法利用序列分解模块将云集群数据分解为趋势项和季节项,趋势项采用双系数网络分别对输入空间预测的均值和方差进行归一化和反归一化,季节项采用融合傅里叶变换的频域自注意力模型进行预测,并在模型训练过程中使用指数移动平均模型动态调整训练损失的误差范围.实验结果表明,对比最先进的五种预测式弹性伸缩算法,本文所提出的方法在保持较低的模型训练和推理时间下,不同预测窗口单变量与多变量预测均方误差分别降低了10.07%和10.01%.
