首页|基于多维数据场可视化算法的数学模型优化系统设计

基于多维数据场可视化算法的数学模型优化系统设计

扫码查看
多维数据场可视化算法能够通过将数据映射到可视化空间中的不同维度和属性,以更好地帮助人们理解数据之间的关系和模式.研究针对牵引供电系统的优化问题,引入了四维可视化算法,并利用供电臂长度的不断调整,得到最佳牵引变电所和分区所位置.结果显示,相比于粒子群算法以及多维尺度变换方法,采用四维可视化算法优化后的解集分布更具均匀性以及多样性.同时,四维可视化算法的IGD值明显低于其余两种算法,其最低IGD值仅为6.002 1×104.此外,基于四维可视化算法的优化设计方案的全线功率损失仅为4.76 MW.说明四维可视化算法能够有效地优化牵引供电系统设计,降低功率损失,使得设计方案更加优化和可行,为牵引供电系统的设计和优化提供了有效的指导.
Design of Mathematical Model Optimization System Based on Multidimensional Data Field Visualization Algorithm
The multidimensional data field visualization algorithm can better help people understand the relationships and patterns between data by mapping data to different dimensions and attributes in the visualization space.A four-dimensional visualization algo-rithm was introduced to optimize the traction power supply system,and the optimal location of traction substations and section posts was obtained by continuously adjusting the length of the power supply arm.The results show that compared to particle swarm optimiza-tion and multidimensional scale transformation methods,the optimized solution set distribution using the four-dimensional visualiza-tion algorithm is more uniform and diverse.Meanwhile,the IGD value of the four-dimensional visualization algorithm is significantly lower than the other two algorithms,with a minimum IGD value of only 6.002 1x 104.In addition,the overall power loss of the opti-mized design scheme based on four-dimensional visualization algorithm is only 4.76MW.The four-dimensional visualization algo-rithm can effectively optimize the design of traction power supply systems,reduce power losses,and make the design scheme more op-timized and feasible.It provides effective guidance for the design and optimization of traction power supply systems.

multidimensional data fieldfour dimensional visualizationmathematical modeltraction power supply system

陈博

展开 >

咸阳师范学院,陕西咸阳 712000

多维数据场 四维可视化 数学模型 牵引供电系统

2019年"青年骨干教师"培养项目咸阳师范学院科研基金项目

XSYGG20190912XSYK095

2024

自动化与仪器仪表
重庆工业自动化仪表研究所,重庆市自动化与仪器仪表学会

自动化与仪器仪表

CSTPCD
影响因子:0.327
ISSN:1001-9227
年,卷(期):2024.(5)
  • 12