The second-order differential equations with perturbation terms to solve variational inequality problem were focused on and discussed the convergence of its solution and the speed of the convergence. Firstly,the Karush-Kuhn-Tucker (KKT) conditions of the original variational inequality problem were equivalently transformed into a system of smoothing equations by using a smoothing complementary function,and it was furtherly equivalent to an unconstrained optimization problem. Secondly,a system of second-order differential equations with perturbation terms was established to solve the final unconstrained optimization problem and discuss the stability of the differential equation system and the speed of convergence under the certain conditions. The convergence and the speed of the convergence for the solution to the original variational inequality problem was discussed. Finally,numerical experiments were given to show the effectiveness of the differential equation method for solving the variational inequality problem.
ArrowK J, HurwiczL.Reduction of constrained maxima to saddle-point problems[M]// Traces and Emergence of Nonlinear Programming.Basel:Birkhäuser,2014:61-80
[2]
HopfieldJ J.Neurons with graded response have collective computational properties like those of two⁃state neurons[J].Proceedings of the National Acade my of Sciences,1984,81(10):3088-3092.
[3]
TankD, HopfieldJ.Simple neural optimization networks: an a/d converter,signal decision circuit,and a linear programming circuit[J].IEEE Transactions on circuits and systems1986,33(5):533-541.
[4]
AntipinA.S.Solving variational inequalities with coupling constraints with the use of differential equations[J].Differential Equations,2000,36(11):1587-1596.
[5]
HeB S, YangH.A neural network model for monotone linear asymmetric variational inequalities[J].IEEE Transactions on Neural Networds,2000,11(1):3-16.
[6]
GaoX B, LiaoL Z, QiL.A novel neural network for variational inequalities with linear and nonlinear constraints[J].IEEE Transactions on Neural Networks,2005,16(6):1305-1317.
[7]
SunJ H, ChenJ S, KoC H.Neural networks for solving second⁃order cone constrained variational inequality problem[J].Compu tational Optimization and Applications,2012,51(2):623-648.
[8]
王莉.变分不等式的微分方程方法与增广Lagrange 方法[D].大连:大连理工大学,2011.
[9]
AttouchH, ChbaniZ, RiahiH.Fast convex optimization via time scaling of damped inertial gradient dynamics[J/OL].Mathematics.(2019-05-24)[2021-03-03].
[10]
AttouchH, CabotA.Asymptotic stabilization of inertial gradient dynamics with time⁃dependent viscosity[J].Journal of Differential Equations,2017,263(9):5412-5458.
[11]
AttouchH, ChbaniZ, PeypouquetJ,et al.Fast convergence of inertial dynamics and algorithms with asymptotic vanishing viscosity[J].Mathematical Programming,2018,168(1):123-175.
[12]
SunJ H, FuW C, AlcantaraJ H,et al.A Neural Network Based on the Metric Projector for Solving SOCCVI Problem[J].IEEE Transactions on Neural Networks and Learning Systems,2021,32(7):2886-2900.
[13]
孙菊贺.锥约束变分不等式问题的数值方法的研究[D].大连:大连理工大学,2008.
[14]
FukushimaM.A relaxed projection method for variational inequalities[J].Mathematical Programming,1986,35(1):58-70.
[15]
SunJ H, ZhangL W.A globally convergent method based on fischer-burmeister operators for solving second-order cone constrained variational inequality problems[J].Computers & Mathematics with Applications,2009,58(10):1936-1946.