传统的配电网故障恢复算法难于同时兼顾恢复过程的快速性和恢复策略的最优化。文章提出一种将启发式搜索算法与优化算法相结合的配电网故障阶段式恢复策略:第一阶段采用启发式搜索方法恢复负荷供电;第二阶段利用优化算法处理过载的负荷...传统的配电网故障恢复算法难于同时兼顾恢复过程的快速性和恢复策略的最优化。文章提出一种将启发式搜索算法与优化算法相结合的配电网故障阶段式恢复策略:第一阶段采用启发式搜索方法恢复负荷供电;第二阶段利用优化算法处理过载的负荷转移;第三阶段按启发式搜索方法处理过载负荷的切除。为实现快速的网络拓扑分析,采用家族树结构表征配电网,并对传统的粒子群优化(particle swarm optimization,PSO)算法与模拟退火(simulated annealing,SA)优化算法进行改进,提出了协同进化算法(co-evolutionary algorithm of PSO and SA,CPSOSA),CPSOSA算法在求解故障恢复数学模型时具有较高的全局寻优能力。算例分析证明了本文所提恢复策略及算法的可行性和高效性。展开更多
针对生化过程软测量建模过程中样本数据可能包含的测量误差对模型性能的影响,提出一种自适应加权最小二乘支持向量机(Adaptive weighted least squares support vector machine,AWLS-SVM)回归的软测量建模方法。该方法基于最小二乘支持...针对生化过程软测量建模过程中样本数据可能包含的测量误差对模型性能的影响,提出一种自适应加权最小二乘支持向量机(Adaptive weighted least squares support vector machine,AWLS-SVM)回归的软测量建模方法。该方法基于最小二乘支持向量机模型,根据样本拟合误差,并结合改进的正态分布赋权规则,自适应地为每个建模样本分配不同的权值,以降低随机误差对模型性能的影响;同时采用混沌差分进化—模拟退火(Chaos differential evolution simulated annealing,CDE-SA)算法对模型参数进行优化选择,以提高模型的泛化能力。仿真实验表明,AWLS-SVM模型的预测精度及鲁棒性能优于最小二乘支持向量机(Least squares support vector machine,LS-SVM)和加权最小二乘支持向量机(Weighted least squares support vector machine,WLS-SVM)。利用Pensim仿真平台的数据,将AWLS-SVM方法用于青霉素发酵过程软测量建模,获得了较好的效果。展开更多
Through systematic analysis and comparison of the common features of SAA, ES and traditional LS (local search) algorithm, a new hybrid strategy of mixing SA, ES with LS, namely HIEA (Hybrid Intelligent Evolutionary Al...Through systematic analysis and comparison of the common features of SAA, ES and traditional LS (local search) algorithm, a new hybrid strategy of mixing SA, ES with LS, namely HIEA (Hybrid Intelligent Evolutionary Algorithm), is proposed in this paper. Viewed as a whole, the hybrid strategy is also an intelligent heuristic searching procedure. But it has some characteristics such as generality, robustness, etc., because it synthesizes advantages of SA, ES and LS, while the shortages of the three methods are overcome. This paper applies Markov chain theory to describe the hybrid strategy mathematically, and proves that the algorithm possesses the global asymptotical convergence and analyzes the performance of HIEA.展开更多
文摘传统的配电网故障恢复算法难于同时兼顾恢复过程的快速性和恢复策略的最优化。文章提出一种将启发式搜索算法与优化算法相结合的配电网故障阶段式恢复策略:第一阶段采用启发式搜索方法恢复负荷供电;第二阶段利用优化算法处理过载的负荷转移;第三阶段按启发式搜索方法处理过载负荷的切除。为实现快速的网络拓扑分析,采用家族树结构表征配电网,并对传统的粒子群优化(particle swarm optimization,PSO)算法与模拟退火(simulated annealing,SA)优化算法进行改进,提出了协同进化算法(co-evolutionary algorithm of PSO and SA,CPSOSA),CPSOSA算法在求解故障恢复数学模型时具有较高的全局寻优能力。算例分析证明了本文所提恢复策略及算法的可行性和高效性。
文摘针对生化过程软测量建模过程中样本数据可能包含的测量误差对模型性能的影响,提出一种自适应加权最小二乘支持向量机(Adaptive weighted least squares support vector machine,AWLS-SVM)回归的软测量建模方法。该方法基于最小二乘支持向量机模型,根据样本拟合误差,并结合改进的正态分布赋权规则,自适应地为每个建模样本分配不同的权值,以降低随机误差对模型性能的影响;同时采用混沌差分进化—模拟退火(Chaos differential evolution simulated annealing,CDE-SA)算法对模型参数进行优化选择,以提高模型的泛化能力。仿真实验表明,AWLS-SVM模型的预测精度及鲁棒性能优于最小二乘支持向量机(Least squares support vector machine,LS-SVM)和加权最小二乘支持向量机(Weighted least squares support vector machine,WLS-SVM)。利用Pensim仿真平台的数据,将AWLS-SVM方法用于青霉素发酵过程软测量建模,获得了较好的效果。
文摘Through systematic analysis and comparison of the common features of SAA, ES and traditional LS (local search) algorithm, a new hybrid strategy of mixing SA, ES with LS, namely HIEA (Hybrid Intelligent Evolutionary Algorithm), is proposed in this paper. Viewed as a whole, the hybrid strategy is also an intelligent heuristic searching procedure. But it has some characteristics such as generality, robustness, etc., because it synthesizes advantages of SA, ES and LS, while the shortages of the three methods are overcome. This paper applies Markov chain theory to describe the hybrid strategy mathematically, and proves that the algorithm possesses the global asymptotical convergence and analyzes the performance of HIEA.
