{"title":"基于改进Patankar—Runge—Kutta方法的贝叶斯优化设计时间步长控制器","authors":"Thomas Izgin, Hendrik Ranocha","doi":"arxiv-2312.01796","DOIUrl":null,"url":null,"abstract":"Modified Patankar--Runge--Kutta (MPRK) methods are linearly implicit time\nintegration schemes developed to preserve positivity and a linear invariant\nsuch as the total mass in chemical reactions. MPRK methods are naturally\nequipped with embedded schemes yielding a local error estimate similar to\nRunge--Kutta pairs. To design good time step size controllers using these error\nestimates, we propose to use Bayesian optimization. In particular, we design a\nnovel objective function that captures important properties such as tolerance\nconvergence and computational stability. We apply our new approach to several\nMPRK schemes and controllers based on digital signal processing, extending\nclassical PI and PID controllers. We demonstrate that the optimization process\nyields controllers that are at least as good as the best controllers chosen\nfrom a wide range of suggestions available for classical explicit and implicit\ntime integration methods.","PeriodicalId":501061,"journal":{"name":"arXiv - CS - Numerical Analysis","volume":" 14","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2023-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Using Bayesian Optimization to Design Time Step Size Controllers with Application to Modified Patankar--Runge--Kutta Methods\",\"authors\":\"Thomas Izgin, Hendrik Ranocha\",\"doi\":\"arxiv-2312.01796\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Modified Patankar--Runge--Kutta (MPRK) methods are linearly implicit time\\nintegration schemes developed to preserve positivity and a linear invariant\\nsuch as the total mass in chemical reactions. MPRK methods are naturally\\nequipped with embedded schemes yielding a local error estimate similar to\\nRunge--Kutta pairs. To design good time step size controllers using these error\\nestimates, we propose to use Bayesian optimization. In particular, we design a\\nnovel objective function that captures important properties such as tolerance\\nconvergence and computational stability. We apply our new approach to several\\nMPRK schemes and controllers based on digital signal processing, extending\\nclassical PI and PID controllers. We demonstrate that the optimization process\\nyields controllers that are at least as good as the best controllers chosen\\nfrom a wide range of suggestions available for classical explicit and implicit\\ntime integration methods.\",\"PeriodicalId\":501061,\"journal\":{\"name\":\"arXiv - CS - Numerical Analysis\",\"volume\":\" 14\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-12-04\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Numerical Analysis\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2312.01796\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Numerical Analysis","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2312.01796","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Using Bayesian Optimization to Design Time Step Size Controllers with Application to Modified Patankar--Runge--Kutta Methods
Modified Patankar--Runge--Kutta (MPRK) methods are linearly implicit time
integration schemes developed to preserve positivity and a linear invariant
such as the total mass in chemical reactions. MPRK methods are naturally
equipped with embedded schemes yielding a local error estimate similar to
Runge--Kutta pairs. To design good time step size controllers using these error
estimates, we propose to use Bayesian optimization. In particular, we design a
novel objective function that captures important properties such as tolerance
convergence and computational stability. We apply our new approach to several
MPRK schemes and controllers based on digital signal processing, extending
classical PI and PID controllers. We demonstrate that the optimization process
yields controllers that are at least as good as the best controllers chosen
from a wide range of suggestions available for classical explicit and implicit
time integration methods.