论文部分内容阅读
针对目前风量分布中依靠经验认识全部近似认为是正态分布的情况,利用概率图分析法给出了判断风量分布的严格定义,且给出了基于对数正态分布的风量分布模型,证明了并非所有风量分布均服从正态分布.为避开失效模式的NP难题,提出网络可靠度的近似算法.该方法以递推分解算法为基础,引入多维正态变量积分公式,计算多失效模型的联合概率,进而计算风网系统的可靠度,大大降低了计算复杂性.对于特别复杂的风网,也可以采用上下界逼近的方法得出近似值,并通过实例分析,验证了本方法的高精度和高效率.
In view of the fact that the current air volume distribution relies on the empirical knowledge of all approximations as a normal distribution, a strict definition of the air volume distribution is given by using the probability map analysis method, and the air volume distribution model based on the logarithm normal distribution is given. In order to avoid the NP problem of failure mode, an approximate algorithm of network reliability is proposed.The method is based on the recursive decomposition algorithm, introduces the integral formula of multi-dimensional normal variable, and calculates the multi-failure model Joint probability, and then calculate the reliability of the wind network system, greatly reducing the computational complexity.For a particularly complex wind network, the approximation method can also be approximated using the upper and lower bounds, and the example analysis shows that the method of high precision And high efficiency.