Tent混沌和模拟退火改进的飞蛾扑火优化算法

针对标准飞蛾扑火优化算法存在的易陷入局部最优陷阱、全局寻优能力不足的问题,借鉴混沌序列、模拟退火算法和遗传算法,提出Tent混沌和模拟退火改进的飞蛾扑火优化算法.首先,通过Tent混沌序列初始化种群,增加种群多样性;然后对当前最优解增加扰动产生新解,并与当前最优解按比例杂交相加,根据模拟退火算法中的Metropolis准则判断是否接受杂交后的新解,最终获得最优解.分别使用复杂高维基准函数和航迹规划问题测试算法性能.其中,6个复杂基准函数寻优测试结果表明,对于10维基准函数,该算法经过约0.25秒收敛到最优值;对于50维基准函数,该算法经过约0.5秒收敛到最优值.与标准飞蛾扑火优化算法和其它智能优化算法相比,该算法能够有效跳出局部最优解,寻优精度更高,收敛速度更快.航迹规划仿真表明,对有4个禁飞区和2个威胁源的空域环境,该算法经过大约100次迭代可以得到最优航迹,与标准飞蛾扑火优化算法相比精度更高,具有实际应用价值.因此,该算法具有更好的寻优性能.

Tent chaos and simulated annealing improved moth-flame optimization algorithm

Aiming at the problem that the moth-flame optimization algorithm is easy to fall into the local optimal trap and the global optimization ability is insufficient, chaotic sequences, simulated annealing algorithm, and genetic algorithm were utilized to propose a tent chaos and simulated annealing improved moth-flame optimization algorithm. First, the tent chaotic sequences were used to initialize the population and increase the diversity of the population. Then the new solution was obtained by adding disturbance to the current optimal solution which was further proportionally hybridized to generate a final new solution. According to the Metropolis criterion in the simulated annealing algorithm, the ultimate solution was finally obtained by determining whether the new hybridized solution was accepted. The performance of the proposed algorithm was tested through complex high-dimensional benchmark functions and trajectory planning problems. The results of six complex benchmark function optimization tests show that for the 10-dimensional benchmark function, the algorithm converged to the optimal value after about 0.25 seconds, and for 50 dimensions, it was about 0.5 seconds. Compared with the standard moth-flame optimization algorithm and other intelligent optimization algorithms, the proposed algorithm could effectively jump out of the local optimal solution in the discussed scenarios, and the accuracy precision and the convergence speed were improved in certain degree. The trajectory planning simulation results show that for an airspace environment with four no-fly zones and two threat sources, the proposed algorithm could obtain the optimal trajectory after about 100 iterations, which is more accurate than the standard moth-flame optimization algorithm. Therefore, it is valuable for engineering applications and has better performance.