Averaging in optimal switching control problems

The averaging in optimal switching control problems is considered under the following two cases: the switching cost does not depend on ∈ and the switching cost vanishes as ∈ tends to zero. The value function of the original fast problem converges locally uniformly to the value function of the averaged problem under both cases. The ways of averaging turn out to be different between both cases. This work is supported by the National Natural Science Foundation of China.