求解并联冷机负荷分配问题的改进FODPSO算法

来源期刊:中南大学学报(自然科学版)2021年第6期

论文作者:赵安军 于军琪 赵泽华 王福 陈时羽

文章页码:1901 - 1915

关键词:负荷分配;并联冷机;分数阶达尔文粒子群优化算法;蒙特卡洛;多重优化;自适应多策略

Key words:load distribution; parallel chiller system; fractional order Darwinian particle swarm optimization algorithm(FODPSO); Monte Carlo; multi-optimization; adaptive multi-strategy

摘    要:针对并联冷机负荷分配问题,以系统总功率最小为优化目标,建立满足系统末端负荷需求的并联冷机负荷分配优化模型,提出一种改进分数阶达尔文粒子群优化(IFODPSO)算法,以每台冷机的部分负荷率为优化变量进行求解,优化并联冷机系统的运行策略以节能。首先,针对基本分数阶达尔文粒子群优化(FODPSO)算法粒子初始化过于分散的问题,提出利用蒙特卡洛方法结合基本算数运算符生成初始种群;其次,针对其在高维优化中难以同时搜寻到每一维最优解的问题,引入多重优化提高算法稳定性并加快收敛速度;第三,针对易陷入局部最优的问题,通过自适应多策略行为使粒子能够根据其适应度选择合适的更新方式,提高了算法的搜索能力;最后,以2个典型的并联冷机系统作为案例验证所提出算法的性能,并与其他现有优化算法的实验结果进行对比。研究结果表明:相比于其他算法,IFODPSO算法在并联冷机负荷分配问题的求解中能够取得更加显著的节能效果,得到更优的运行策略,同时收敛精度、收敛速度和稳定性都有了显著提高。

Abstract: Aiming at load distribution problem of parallel chillers, an optimization model of load distribution of parallel chillers was established and the end load requirement was satisfied. An improved fractional order Darwinian particle swarm optimization(IFODPSO) algorithm was proposed to optimize the operation strategy with the partial load rate of each chiller as the optimization variable to save energy. Firstly, Monte Carlo method combined with the basic arithmetic operator was used to solve the problem of too much scattered particle initialization in the basic fractional order Darwinian particle swarm optimization(FODPSO) algorithm. Secondly, multi-optimization was introduced to find the optimal solution of each dimension at the same time, which improved the stability and accelerated the convergence speed of the algorithm. Thirdly, the adaptive multi-strategy behavior enabled particles to choose appropriate update mode according to its fitness to improve the search ability of the algorithm for local optimal problem. Finally, two typical parallel chiller systems were taken as examples to verify the performance of the proposed algorithm, and the experimental results of other existing optimization algorithms were compared. The results show that compared with other algorithms, IFODPSO algorithm can achieve more significant energy-saving effect in solving the load distribution problem of parallel chillers, and can search for a better operation strategy. Meanwhile, the convergence accuracy, convergence speed and stability are significantly improved.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号