为解决气动调节阀控制过程中出现的超调大、精度低等问题,本文采用BP神经网络整定出较优的PID(Proportional Integral Derivative)控制参数,对Smith预估控制器以及模糊控制器进行设计,实现了基于BP神经网络的Smith-Fuzzy-PID控制方法。...为解决气动调节阀控制过程中出现的超调大、精度低等问题,本文采用BP神经网络整定出较优的PID(Proportional Integral Derivative)控制参数,对Smith预估控制器以及模糊控制器进行设计,实现了基于BP神经网络的Smith-Fuzzy-PID控制方法。搭建了实验平台,通过阶跃响应实验来对控制方法进行验证,验证结果表明,提出的方法调节过程无超调,调节时间仅为1.9 s,定位精度在±0.5%以内,有效提高了系统的稳定性,实现了气动调节阀的快速精准定位。展开更多
The heterogeneity in the communicating terminals needs to he handled through software-supported adaptation, The real-time dynamical adaptation brings the system more time cost unavoidably. In this paper, we propose Qo...The heterogeneity in the communicating terminals needs to he handled through software-supported adaptation, The real-time dynamical adaptation brings the system more time cost unavoidably. In this paper, we propose QoS pre-estimation for user admission control and system load control. Feedback acts as adjuster of the input parameters of the pre-estimate module. We present tests results to evaluate our approach, The test results show that our service quality control mechanism is effective.展开更多
The heterogeneity in the communicating terminals needs to be handled through software-supported adaptation. A QoS （quality of service） pre-estimation for user admission control and system load control is proposed. I...The heterogeneity in the communicating terminals needs to be handled through software-supported adaptation. A QoS （quality of service） pre-estimation for user admission control and system load control is proposed. In the QoS pre-estimation, the content quality and wait time characteristics are combined together. The feedback control that acts as an adjuster of the input parameters of the pre-estimate module to improve the estimate accuracy is also described in detail. Test results show that the service quality control mechanism is effective, and the system capacity can be improved.展开更多
光伏电网频率调整过程中,依靠常规Smith预估控制器实现电网调频控制,对模型精度具有较强的依赖性,控制策略实施后最大频率变化率(rate of change of frequency,RoCoF)较大。因此,提出基于改进型Smith预估计器与大数据的光伏电网调频逐...光伏电网频率调整过程中,依靠常规Smith预估控制器实现电网调频控制,对模型精度具有较强的依赖性,控制策略实施后最大频率变化率(rate of change of frequency,RoCoF)较大。因此,提出基于改进型Smith预估计器与大数据的光伏电网调频逐步惯性控制方法。首先,采集历史气象数据和光伏电网运行数据,应用大数据分析领域的密度峰值聚类算法进行划分处理,再筛选相似日数据输入长短期记忆网络中,预测出未来光伏发电的功率变化;然后,依托逐步惯性控制思想,设计包含短时超发、转速恢复等多个阶段的电网调频控制策略,将模糊自适应比例-积分-微分(proportion-integration-differentiation,PID)控制器融入常规Smith预估计器,从而升级得到优化版的Smith预估计器;最后,在不受被控模型变化影响的情况下,依据预估补偿原理完成逐步惯性调频控制,并应用麻雀搜索算法求解出最优控制参数。实验结果表明:该控制方法实施后,光伏电网运行过程中最大RoCoF仅为0.086 Hz/s,有效降低了对模型精度的依赖性,保证了电力系统的稳定运行。展开更多
文摘为解决气动调节阀控制过程中出现的超调大、精度低等问题,本文采用BP神经网络整定出较优的PID(Proportional Integral Derivative)控制参数,对Smith预估控制器以及模糊控制器进行设计,实现了基于BP神经网络的Smith-Fuzzy-PID控制方法。搭建了实验平台,通过阶跃响应实验来对控制方法进行验证,验证结果表明,提出的方法调节过程无超调,调节时间仅为1.9 s,定位精度在±0.5%以内,有效提高了系统的稳定性,实现了气动调节阀的快速精准定位。
基金Supported by the National Natural Science Foun-dation of China (60172075 ,60472050)
文摘The heterogeneity in the communicating terminals needs to he handled through software-supported adaptation, The real-time dynamical adaptation brings the system more time cost unavoidably. In this paper, we propose QoS pre-estimation for user admission control and system load control. Feedback acts as adjuster of the input parameters of the pre-estimate module. We present tests results to evaluate our approach, The test results show that our service quality control mechanism is effective.
基金The National Natural Science Foundationof China (No. 60172075 and No.60472050)
文摘The heterogeneity in the communicating terminals needs to be handled through software-supported adaptation. A QoS （quality of service） pre-estimation for user admission control and system load control is proposed. In the QoS pre-estimation, the content quality and wait time characteristics are combined together. The feedback control that acts as an adjuster of the input parameters of the pre-estimate module to improve the estimate accuracy is also described in detail. Test results show that the service quality control mechanism is effective, and the system capacity can be improved.
文摘光伏电网频率调整过程中,依靠常规Smith预估控制器实现电网调频控制,对模型精度具有较强的依赖性,控制策略实施后最大频率变化率(rate of change of frequency,RoCoF)较大。因此,提出基于改进型Smith预估计器与大数据的光伏电网调频逐步惯性控制方法。首先,采集历史气象数据和光伏电网运行数据,应用大数据分析领域的密度峰值聚类算法进行划分处理,再筛选相似日数据输入长短期记忆网络中,预测出未来光伏发电的功率变化;然后,依托逐步惯性控制思想,设计包含短时超发、转速恢复等多个阶段的电网调频控制策略,将模糊自适应比例-积分-微分(proportion-integration-differentiation,PID)控制器融入常规Smith预估计器,从而升级得到优化版的Smith预估计器;最后,在不受被控模型变化影响的情况下,依据预估补偿原理完成逐步惯性调频控制,并应用麻雀搜索算法求解出最优控制参数。实验结果表明:该控制方法实施后,光伏电网运行过程中最大RoCoF仅为0.086 Hz/s,有效降低了对模型精度的依赖性,保证了电力系统的稳定运行。
