中文摘要：

针对入侵杂草优化算法收敛速度较慢、易陷入局部最优的缺点,提出了一种改进的入侵杂草优化算法。首先,采用反向学习初始化方法对种群进行初始化以提高其全局收敛速度;其次,利用改进蜂群算法中的全局引导搜索策略,对繁殖后的种子进行最优引导搜索以提高其跳出局部最优点的能力。最后,对不同维数的5个标准测试函数进行了仿真验证。试验结果表明：与GABC及标准IWO（Invasive Weed Optimization）算法相比较,该改进算法在函数优化方面具有较快的收敛速度和较强的跳出局部最优的能力。

英文摘要：

Aiming at the problems of invasive weed optimisation（ IWO） algorithm which is slow in convergence rate and easy to fall into local optimum,we propose an improved IWO algorithm. First,we initialise the population with reverse learning initialisation method to improveits global convergence rate; Secondly,we use global guide search stagey of the improved bee colony algorithm to conduct optimal guide search of the bred seeds in order to enhance their capability of jumping out the local optima. Finally,we carry out simulation verificationson 5 benchmark functions with different dimensions. Test results show that the improved algorithm proposed has faster convergence rate and stronger capability in jumping out local optimum in the aspect of function optimisation than the GABC and standard IWO algorithm.