论文部分内容阅读
针对标准遗传算法采用二进制编码离散连续函数时存在映射误差和不便于确定特定知识的问题,提出了基于实码加速的遗传算法优化收敛的投影寻踪灌排模式评价模型,克服了容易陷入局部最优的缺点,提高了全局搜索能力和收敛速度。以江西灌排试验站2年种植水稻的实测数据为例,选用经济效益指标、环境效益指标和社会效益指标综合计算投影值,基于实码加速的投影寻踪模型对赣抚平原地区不同灌排模式进行评价。结果表明,该模型有效解决了标准遗传算法的寻优效率依赖于优化变量初始变化区间的大小的缺点,将它用于优化选择灌排模式评价是切实可行的,对农业生产实践起到了一定的指导作用。
In order to solve the problem of mapping error and inconvenient identification of specific knowledge when binary genetic algorithm adopts discrete continuous function of binary code, a genetic algorithm based on real code optimization is proposed to evaluate the convergence of projection-based model. Excellent shortcomings, improved global search capabilities and convergence speed. Taking the measured data of 2 years of rice cultivation in Jiangxi irrigation and drainage experimental station as an example, the projections of economic benefits, environmental benefits and social benefits were selected. Based on the projection-pursuit model with real code acceleration, Row mode evaluation. The results show that the model effectively solves the shortcoming of the optimization efficiency of standard genetic algorithm depends on the size of the initial variation interval of the optimization variables. It is feasible to use it in the evaluation of the optimal selection of irrigation and drainage patterns, which has certain effect on the agricultural production practice Guidance.