Voltage quality impacts the stability, efficiency, and cost effectiveness of the distribution network. With the integration of large-scale wind power and solar photovoltaic power generation, the probability of fluctuations in the feeder network voltage of the distribution network is greatly increased, and the study of voltage stability is more important. This paper proposes a comprehensive scheme to optimize voltage quality problems from the network level instead of the traditional scheme from the point level. The effect of different compensation points on the voltage of distribution network is analyzed by simulation. Then the elite selection strategy of genetic algorithm is used to calculate the optimal location of reactive compensators in distribution network. The analysis of IEEE 18-node distribution system simulation shows that the proposed comprehensive control method can generate low voltage deviation with fewer control points, compared with the full compensation cases.

The dual active bridge (DAB) converter is a popular topology with the ability of galvanic isolation and bidirectional power flow. In the practical application, there could be the significant transient current overshoot between the different operating conditions or in the start-up phase of the DAB converter. Therefore, a discrete-time model based model predictive control method is proposed in this paper to constrain the transient inductor current overshoot. An experimental prototype is also built to validate the proposed method. The experimental results are consistent with simulation waveforms and the transient current overshoot can also be effectively constrained to mitigate its stress on the switch devices.

Voltage quality is an important aspect of power quality and a stable voltage amplitude is significant for network security and operational economy. Reactive voltage compensation is mostly paid attention on the transmission grid side, while the reactive voltage on the distribution grid side is rarely noticed. Compensating the reactive voltage of each node is uneconomical, so using a small amount of treatment equipment to meet the voltage requirements of each node of the distribution network is quite necessary. In this paper, a genetic algorithm-based method is developed for positioning of reactive voltage compensator. The effectiveness of the algorithm is verified by a 4-node system. The results indicate that the proposed genetic algorithm-based method can effectively calculate the position of the compensator.

The traditional control strategies of shunt active power filter (SAPF) is primarily focused on the compensation of local non-linear loads, which may become uneconomical for the network with a number of distributed non-linear loads. In this paper, two methods applied to the allocation and sizing of multiple shunt active filters in distribution network are proposed and compared, including an improved particle swarm optimization (PSO) algorithm with one mixed objective function and a multi-objective particle swarm optimization (MOPSO) algorithm. To evaluate the capability of the proposed methods, the IEEE 18-bus test system is employed in simulation. Simulation results confirms that both methods can achieve the goal but the MOPSO-based algorithm is more efficient and universal in the allocation and sizing of multiple SAPFs compared with the PSO based algorithm.