Algorithms for finding global and local equilibrium points of Nash-Cournot equilibrium models involving concave cost

Le Dung Muu,Nguyen Van Quy

DOI: https://doi.org/10.48550/arXiv.1805.02171

2018-05-06

Optimization and Control

Abstract:We consider Nash-Cournot oligopolistic equilibrium models involving separable concave cost functions. In contrast to the models with linear and convex cost functions, in these models a local equilibrium point may not be a global one. We propose algorithms for finding global and local equilibrium points for the models having separable concave cost functions. The proposed algorithms use the convex envelope of a separable concave cost function over boxes to approximate a concave cost model with an affine cost one. The latter is equivalent to a strongly convex quadratic program that can be solved efficiently. To obtain better approximate solutions the algorithms use an adaptive rectangular bisection which is performed only in the space of concave variables Computational results on a lot number of randomly generated data show that the proposed algorithm for global equilibrium point are efficient for the models with moderate number of concave cost functions while the algorithm for local equilibrium point can solve efficiently the models with much larger size.