Xiaoshun Zhang

Accelerating bio-inspired optimizer with transfer reinforcement learning for reactive power optimization

This paper proposes a novel accelerating bio-inspired optimizer (ABO) associated with transfer reinforcement learning (TRL) to solve the reactive power optimization (RPO) in large-scale power systems. A memory matrix is employed to represent the memory of different state-action pairs, which is used for knowledge learning, storage, and transfer among different optimization tasks. Then an associative memory is introduced to significantly reduce the dimension of memory matrix, in which more than one element can be simultaneously updated by the cooperating multi-bion. The win or learn fast policy hill-climbing (WoLF-PHC) is also used to accelerate the convergence. Thus, ABO can rapidly seek the closest solution to the exact global optimum by exploiting the prior knowledge of the source tasks according to their similarities. The performance of ABO has been evaluated for RPO on IEEE 118-bus system and IEEE 300-bus system, respectively. Simulation results verify that ABO outperforms the existing artificial intelligence algorithms in terms of global convergence ability and stability, which can raise one order of magnitude of the convergence rate than that of others.

Hierarchically correlated equilibrium Q-learning for multi-area decentralized collaborative reactive power optimization

A hierarchically correlated equilibrium Q-learning (HCEQ) algorithm for reactive power optimization that considers carbon emission on the grid-side as an optimization objective, is proposed here. Based on the multi-area decentralized collaborative framework, the controllable variables in each region are divided into several optimization layers, which is an effective method for solving the limitations posed by dimensionality. The HCEQ provides constant information on the interaction between the state-action value function matrices, as well as on the cooperative game equilibrium among agents in each region. After acquiring the optimal value function matrix in the pre-learning process, HCEQ is able to quickly achieve an optimal solution online. Simulation of the IEEE 57-bus system is performed, which demonstrates that the proposed algorithm can effectively solve multi-area decentralized collaborative reactive power optimization, with the desired global search capabilities and convergence speed.

Optimal power tracking of doubly fed induction generator-based wind turbine using swarm moth–flame optimizer:

This paper proposes a novel swarm moth–flame optimizer (SMFO) to obtain the optimal parameters of four interacting proportional–integral (PI) loops of a doubly fed induction generator (DFIG)-based wind turbine, so that maximum power point tracking (MPPT) may be achieved together with an improved fault ride-through (FRT) capability. The SMFO is inspired by a moth swarm encircling a flame at night, in which each flame is simultaneously encircled by multiple moths for a greater exploitation, whereas the flame with a higher brightness (i.e. a smaller fitness function) will attract more moths than those of its adjacent flames. In order to achieve a wider exploration, a ring network is then constructed among the flames such that the moths may be guided to search for a brighter flame more effectively. Three case studies are undertaken, which verify that an improved global convergence, more optimal power tracking and enhanced FRT capability may be achieved by SMFO compared with those of existing meta-heuristic techniques.