为提高超短期风速预测的可靠性和准确性,将被预测地点(本地)周边测风塔风速风向等当前和最近历史观测值作为基础数据,采用空间相关性来预测本地的未来风速。首先,依据风向和风速的延迟相关性,挑选出上游测风塔。之后,结合最优延迟时间,...为提高超短期风速预测的可靠性和准确性,将被预测地点(本地)周边测风塔风速风向等当前和最近历史观测值作为基础数据,采用空间相关性来预测本地的未来风速。首先,依据风向和风速的延迟相关性,挑选出上游测风塔。之后,结合最优延迟时间,利用各上游和本地最近的风速观测值来训练预测模型。最后,将各上游风速的当前观测值输入模型,即可得到本地的风速预测值。以偏最小二乘回归(partial least squares regression,PLSR)为主要模型,并采用线性回归(linear regression,LR)、最小二乘支持向量回归等模型进行对照。以冬季风时期的荷兰Huibertgat和天津为被预测地点,进行了PLSR、LR预测误差与模型阶数、样本容量之间关系的数值实验。研究表明,在冬季风时期,当样本容量达到一定程度后,预测误差的变化对阶数、样本容量和模型的类型均不再敏感。这表明空间相关性是一种较为可靠的超短期风速预测方法。展开更多
受湍流影响,室内通风环境下的烟羽分布表现出波动变化且不连续的特性;在一些角落处,较大的漩涡会产生长时间的局部浓度极值区;另外室内的障碍物也会改变烟羽的分布状况.因此室内有障碍通风环境下的机器人气味源搜索问题变得很复杂.本文...受湍流影响,室内通风环境下的烟羽分布表现出波动变化且不连续的特性;在一些角落处,较大的漩涡会产生长时间的局部浓度极值区;另外室内的障碍物也会改变烟羽的分布状况.因此室内有障碍通风环境下的机器人气味源搜索问题变得很复杂.本文提出了基于概率适应度函数的粒子群优化(Probability-fitness-function based particle swarm optimization,P-PSO)算法并用于多机器人气味源搜索.P-PSO算法的特点是采用概率而非确定数来表达适应度函数值.针对气味源搜索问题,P-PSO算法的适应度函数值由贝叶斯和变论域模糊推理估计的气味源概率表达.为验证提出的搜索策略,构建了对应实际边界条件的室内通风环境的烟羽模型.仿真研究证明了本文提出的P-PSO搜索算法用于解决气味源搜索问题的可行性.展开更多
Oil holdup of oil-water two-phase flow was measured by using platinum resistance based on the fluid thermal balance equation.In order to improve the measurement accuracy of oil holdup,the effects of the electrical hea...Oil holdup of oil-water two-phase flow was measured by using platinum resistance based on the fluid thermal balance equation.In order to improve the measurement accuracy of oil holdup,the effects of the electrical heater fore-and-aft temperature difference of platinum resistance and total oil-water flux on oil holdup were researched.A least squares support vector machine(LSSVM)model with parameters optimized by genetic algorithm(GA)was proposed,the temperature difference and total flux of oil-water two-phase flow were used as inputs,and the oil holdup was used as output of the LSSVM model and the ideal model of oil holdups was obtained.The oil holdup model based on least squares support vector machine and genetic algorithm(LSSVM-GA) was compared with the theory corrected model and good oil holdup measurement results were obtained.The average measurement error was 0.96% in the range of 5% to 60% oil holdup.展开更多
文摘为提高超短期风速预测的可靠性和准确性,将被预测地点(本地)周边测风塔风速风向等当前和最近历史观测值作为基础数据,采用空间相关性来预测本地的未来风速。首先,依据风向和风速的延迟相关性,挑选出上游测风塔。之后,结合最优延迟时间,利用各上游和本地最近的风速观测值来训练预测模型。最后,将各上游风速的当前观测值输入模型,即可得到本地的风速预测值。以偏最小二乘回归(partial least squares regression,PLSR)为主要模型,并采用线性回归(linear regression,LR)、最小二乘支持向量回归等模型进行对照。以冬季风时期的荷兰Huibertgat和天津为被预测地点,进行了PLSR、LR预测误差与模型阶数、样本容量之间关系的数值实验。研究表明,在冬季风时期,当样本容量达到一定程度后,预测误差的变化对阶数、样本容量和模型的类型均不再敏感。这表明空间相关性是一种较为可靠的超短期风速预测方法。
文摘受湍流影响,室内通风环境下的烟羽分布表现出波动变化且不连续的特性;在一些角落处,较大的漩涡会产生长时间的局部浓度极值区;另外室内的障碍物也会改变烟羽的分布状况.因此室内有障碍通风环境下的机器人气味源搜索问题变得很复杂.本文提出了基于概率适应度函数的粒子群优化(Probability-fitness-function based particle swarm optimization,P-PSO)算法并用于多机器人气味源搜索.P-PSO算法的特点是采用概率而非确定数来表达适应度函数值.针对气味源搜索问题,P-PSO算法的适应度函数值由贝叶斯和变论域模糊推理估计的气味源概率表达.为验证提出的搜索策略,构建了对应实际边界条件的室内通风环境的烟羽模型.仿真研究证明了本文提出的P-PSO搜索算法用于解决气味源搜索问题的可行性.
文摘Oil holdup of oil-water two-phase flow was measured by using platinum resistance based on the fluid thermal balance equation.In order to improve the measurement accuracy of oil holdup,the effects of the electrical heater fore-and-aft temperature difference of platinum resistance and total oil-water flux on oil holdup were researched.A least squares support vector machine(LSSVM)model with parameters optimized by genetic algorithm(GA)was proposed,the temperature difference and total flux of oil-water two-phase flow were used as inputs,and the oil holdup was used as output of the LSSVM model and the ideal model of oil holdups was obtained.The oil holdup model based on least squares support vector machine and genetic algorithm(LSSVM-GA) was compared with the theory corrected model and good oil holdup measurement results were obtained.The average measurement error was 0.96% in the range of 5% to 60% oil holdup.
