基于改进蚁群算法的“N-1”安全输电网优化规划

Transmission Network Expansion Planning with N-1 Security Criterion Based on Improved Ant Colony Algorithm

满足N-1安全准则的输电网络扩展规划是一个复杂的非线性组合优化问题。蚁群算法是一种解决组合最优化问题的高效的启发式方法,但容易出现停滞和陷入局部最优。为此,结合满足N-1安全准则的输电网络规划问题的特点,用罚函数方法建立了网络扩展规划性能指标。通过改变蚁群算法中信息增量的选择方法,有效地减小了算法的全局搜索能力和计算效率之间的矛盾。该方法不需初始可行解,在提高计算效率的同时加大了找到全局最优解的概率。通过对IEEE-6节点和IEEE-24节点两个算例的测试,验证了这种方法能有效地解决直接求解满足N-1安全检验的输电网最优扩展规划问题。

Transmission network expansion planning with security considered is a complex nonlinear combinatorial optimization problem. Ant colony algorithm（ACA） is a highly efficient meta-heuristic method for solving combinatorial optimization problem. However, it is inclined to premature convergence problems. In this paper, the DC model is used for the electrical network, and the way of selecting the increment of information in ACA is modified, which reduces the computation complexity, and ensures the global convergency as well. The improved ACA is introduced to directly obtain the global optimal solution for the transmission expansion planning with N--1 security constraints. The proposed method avoids the problem of non-optimal results in two phases method, and nolonger requires the initial feasible solutions in the genetic algorithm. The results of the IEEE 6-bus and IEEE 24-bus systems show the excellent performance of the proposed methodology.