The use of communication networks in control loops has gained increasing attention in recent years due to its advantages and flexible applications. The network quality-of-service （QoS） in those socalled networked co...The use of communication networks in control loops has gained increasing attention in recent years due to its advantages and flexible applications. The network quality-of-service （QoS） in those socalled networked control systems always fluctuates due to changes of the traffic load and available network resources, This paper presents an intelligent scheduling controller design approach for a class of NCSs to handle network QoS variations, The sampling period and control parameters in the controller are simultaneously scheduled to compensate for the network QoS variations. The estimation of distribution algorithm is used to optimize the sampling period and control parameters for better performance. Compared with existing networked control methods, the controller has better ability to compensate for the network QoS variations and to balance network loads. Simulation results show that the plant setting time with the intelligent scheduling controller is reduced by about 64.0% for the medium network load and 49.1% for high network load and demonstrate the effectiveness of the proposed approaches.展开更多
构造了一种基于Alopex(Algorithm of pattern extraction)和分布估计算法(Estimation of distribution algorithm,EDA)相融合的进化算法EDA-Alopex。该算法将分布估计算法嵌入到一种基于Alopex的群智能进化算法(Alopex-based evolutiona...构造了一种基于Alopex(Algorithm of pattern extraction)和分布估计算法(Estimation of distribution algorithm,EDA)相融合的进化算法EDA-Alopex。该算法将分布估计算法嵌入到一种基于Alopex的群智能进化算法(Alopex-based evolutionary algorithm,AEA)中,利用分布估计算法收敛速度快及与传统进化算法进化模式不同的特点来改进AEA算法。新算法综合了AEA算法搜索得到的个体间相关性信息和EDA搜索过程中得到的全局概率信息,能够更好地指导种群向有利的区域进化。仿真结果表明:EDA改进的EDA-Alopex算法搜索性能与AEA算法的搜索性能相比有较大提高,特别是其收敛速度与AEA算法相比有明显提高。展开更多
本文构造了基于分布估计算法(Estimation of Distribution Algorithm,EDA)和遗传算法(GeneticAlgorithm,GA)融合的神经网络(Neural Network,NN)故障诊断模型。传统的GA看作是对生物进化"微观"层面上的模拟,则EDA是对生物进化&...本文构造了基于分布估计算法(Estimation of Distribution Algorithm,EDA)和遗传算法(GeneticAlgorithm,GA)融合的神经网络(Neural Network,NN)故障诊断模型。传统的GA看作是对生物进化"微观"层面上的模拟,则EDA是对生物进化"宏观"层面上的建模,是一种全新的进化模式。EDA与GA融合的实质是在解空间"宏观"和"微观"两个层面进行寻优,可克服NN陷入局部最小,提高NN的泛化能力,使故障诊断的容错性能得到有效改善。将该模型用于高压输电线系统的故障诊断,并作容错性能的评估。由仿真测试表明,研究模型的容错性能要优于传统的BP-NN模型和单纯GA优化NN模型。因此,新诊断模型是有一定的理论和实用价值的。展开更多
基金the National Key Basic Research and Development Program (973) of China (No. 2002cb312205)the National Natural Science Foundation for Key Technical Research of China (No. 60334020)the National Natural Science Foundation of China (Nos. 60574035 and 60674053)
文摘The use of communication networks in control loops has gained increasing attention in recent years due to its advantages and flexible applications. The network quality-of-service （QoS） in those socalled networked control systems always fluctuates due to changes of the traffic load and available network resources, This paper presents an intelligent scheduling controller design approach for a class of NCSs to handle network QoS variations, The sampling period and control parameters in the controller are simultaneously scheduled to compensate for the network QoS variations. The estimation of distribution algorithm is used to optimize the sampling period and control parameters for better performance. Compared with existing networked control methods, the controller has better ability to compensate for the network QoS variations and to balance network loads. Simulation results show that the plant setting time with the intelligent scheduling controller is reduced by about 64.0% for the medium network load and 49.1% for high network load and demonstrate the effectiveness of the proposed approaches.
文摘构造了一种基于Alopex(Algorithm of pattern extraction)和分布估计算法(Estimation of distribution algorithm,EDA)相融合的进化算法EDA-Alopex。该算法将分布估计算法嵌入到一种基于Alopex的群智能进化算法(Alopex-based evolutionary algorithm,AEA)中,利用分布估计算法收敛速度快及与传统进化算法进化模式不同的特点来改进AEA算法。新算法综合了AEA算法搜索得到的个体间相关性信息和EDA搜索过程中得到的全局概率信息,能够更好地指导种群向有利的区域进化。仿真结果表明:EDA改进的EDA-Alopex算法搜索性能与AEA算法的搜索性能相比有较大提高,特别是其收敛速度与AEA算法相比有明显提高。
文摘本文构造了基于分布估计算法(Estimation of Distribution Algorithm,EDA)和遗传算法(GeneticAlgorithm,GA)融合的神经网络(Neural Network,NN)故障诊断模型。传统的GA看作是对生物进化"微观"层面上的模拟,则EDA是对生物进化"宏观"层面上的建模,是一种全新的进化模式。EDA与GA融合的实质是在解空间"宏观"和"微观"两个层面进行寻优,可克服NN陷入局部最小,提高NN的泛化能力,使故障诊断的容错性能得到有效改善。将该模型用于高压输电线系统的故障诊断,并作容错性能的评估。由仿真测试表明,研究模型的容错性能要优于传统的BP-NN模型和单纯GA优化NN模型。因此,新诊断模型是有一定的理论和实用价值的。
