针对扩展蚁群算法收敛慢,且容易陷入局部最优的缺点对扩展蚁群算法提出改进策略.引入量子比特表示蚂蚁位置以增加解的多样性；采用量子非门实现蚂蚁位置的变异以避免蚂蚁陷入局部最优；引入量子旋转门和高斯核概率密度函数结合更新蚂蚁携带的量子比特,利于在连续空间寻优；根据解的重要性改进解存储器中每个解的权值以提高解的方向性,快速获得最优解.通过对多个二维和多维连续函数的对比仿真实验验证了算法的有效性.%According to extension of ant colony optimization converging slowly and easily falling into the local opti-mum, some improved strategies were presented. First, a group of quantum bits which represent the position of the ant increased the diversity of the solution. Secondly, some quantum bits were mutated by a quantum non-gate so as to avoid ants involving in the local optimum. Thirdly, quantum bits of the ant were updated by quantum rotation gates and Gaussian kernel functions so as to find the optimization in continuous space. Lastly, in order to obtain the optimal solution quickly and increase the direction, the weight of each solution in the memory was improved by the importance of the weight. The simulation and comparison on the many two-dimensional and multi-dimensional con-tinuous functions prove the effectiveness of the algorithm.
展开▼
