The smart grid concept envisages the creation of a developed automatic control system of the electric power system operation that involves a wide application of distributed generation plants integrated into the energy clusters. These technologies are fully applicable to the railway electricity supply systems (RESS) which are mainly intended for the reliable high quality electricity supply of the traction consumers (trains) as well as non-traction and non-transport ones. For effective operation of the distributed generation plants, it is necessary to solve the problem of the optimal tuning automatic excitation regulators (AER) and automatic speed regulators (ASR), taking into account the specific features of RESS. The paper presents a description of an adaptive genetic algorithm for solving the problems of settings optimization for AER and ASR of the distributed plant turbine generators operating in RESS. The studies of computer models of electric power systems with distributed generation plants in MATLAB show that the AER and ASR tuning coefficients calculated with the proposed adaptive genetic algorithm provide the required stability margin, good damping electromechanical oscillations and standard power quality values in the power supply system of non-traction consumers.

A trend associated with distributed generator usage is currently being developed. Distributed generators allow to unload the main grid and enhance reliability of consumer power supply, and electric energy quality. In this case, the problem of finding the optimal settings of automatic field / speed regulators shall be solved to provide reliable operation and integration of distributed generators made of synchronous turbo and hydro-generators in power supply systems. This requires usage of large-scale power supply systems and laborious calculations of a large number of interconnected parameters. The other approach providing regulator adjustment by one parameter is possible as well. In this regard, it is proposed to transfer from the adjustment by a current error value to the control implying prognosis calculation. A regulator, which controls a device by a prognosis value, is called as prognostic. The article describes an automatic prognostic speed regulator of a distributed synchronous generator; the influence of an automatic prognostic regulator on frequency control quality in a power supply system, while its operation mode varies, is studied. Results of computer modelling show that usage of an automatic prognostic speed regulator under disturbances of power supply systems allow to enhance quality of frequency control in a power supply system containing distributed generators. The proposed automatic prognostic speed regulator can be used to enhance reliability and integration of distributed generator power supply systems.