Firefly algorithm (FA) is a meta-heuristic optimization algorithm inspired by nature. Due to its superior performance, it has been widely used in real life. However, it also has some shortcomings in some optimization cases, such as low solution accuracy and slow solution speed. Therefore, in this paper, distributed parallel firefly algorithm (DPFA) with four communication strategies is presented to improve these shortcomings. The distributed parallel technique is implanted to divide the initial fireflies into several subgroups, and exchange the information based on communication strategies among subgroups after the fixed iteration. The communication strategies include the maximum of the same group, the average of the same group, the maximum of different groups and the average of different groups. For verifying its performance, this paper compared DPFA with famous optimization algorithms, and experimental results show that DPFA has stronger competitiveness under the test suite of CEC2013. Furthermore, the proposed DPFA is also applied to the PID parameter tuning of variable pitch wind turbine, and conducted experiments show that DPFA outperforms other algorithms. It can smooth the power output and reduce the impact on the power grid when the wind speed fluctuates.

萤火虫算法（FA）是一种受自然启发的元启发式优化算法。由于其优越的性能，在现实生活中得到了广泛的应用。但是，它在一些优化情况下也存在一些缺点，如求解精度低、求解速度慢。因此，本文提出了具有四种通信策略的分布式并行萤火虫算法（DPFA）来改善这些缺点。植入分布式并行技术，将初始萤火虫分成若干子群，固定迭代后子群间基于通信策略交换信息。传播策略包括同组的最大值、同组的平均值、不同组的最大值和不同组的平均值。为了验证其性能，本文将DPFA与著名的优化算法进行了对比，实验结果表明DPFA在CEC2013的测试套件下具有更强的竞争力。此外，所提出的 DPFA 还应用于变桨风力发电机的 PID 参数整定，并且进行的实验表明 DPFA 优于其他算法。可在风速波动时平滑功率输出，减少对电网的冲击。