基于遗传编码的光伏MPPT模式搜索跟踪控制

摘要

In order to solve the maximum power point tracking (MPPT) problem in photovoltaic system, a new MPPT method based on both genetic encoding and pattern search was proposed. The genetic encoding has excellent flexibility, randomicity and parallelism, and can not be influenced by the continuously differentiable performance of fitness function. The globle search range can be adjusted quickly due to the self-adaptability of the genetic algorithm when the environment conditions change. Based on the global search, the local search was enhanced by the pattern search in order to improve the search performance from search efficiency and precision aspects. The simulated and experimental results show that the proposed method can overcome the power loss caused by MPP oscillating phenomenon and improve the tracking efficiency, compared with the traditional hill climbing method and single genetic algorithm control. When the environment conditions change, the tracking effect is still efficient, stable and accurate.%为解决当前太阳能光伏发电系统中最大功率点跟踪(MPPT)问题,采用遗传算法编码技术与模式搜索相结合的MPPT算法,利用遗传编码灵活、随机、采用并行搜索机制且不受适应度函数连续可微影响的特点进行粗搜索.当外界条件发生变化时,其自适应的特点使得粗搜索的范围迅速调整,在此基础上采用模式搜索增强细搜索功能,旨在从速度和精度两方面提高算法搜索性能.仿真及实验表明,与传统爬山法和单一遗传算法控制相比,该方法不但能够克服MPP振荡带来的功率损失,而且能够提高跟踪效率,当外界条件改变时,仍然能够表现出快速稳定而且准确的跟踪效果.