Comparison of PSO and de in the task of optimal control of chaotic Lozi map