论文部分内容阅读
针对现有基于云模型的相关性度量方法缺少必要约束条件的问题,提出一种基于含熵期望曲线的云模型相关性度量方法.将云模型中超熵期望曲线的与区域和或区域的面积比作为相关性的度量基准,解决云模型的区间约束以及半云度量问题.利用3熵(3σ)约束增加云滴的聚集,减少计算开销.将超熵纳入计算,考虑云滴厚度对云模型相关性的影响.本方法克服了面向随机云滴的距离度量方法和数字特征变换方法中存在的计算复杂度高、结果不稳定的问题,同时满足了三类约束的实际计算需求.实验表明该方法能够客观有效地对云模型相关性进行度量,并在基于云模型的系统评价任务中得到了验证.
Aiming at the lack of the necessary constraints for the existing cloud-based model, a new cloud model correlation measure based on entropy-containing expectation curve is proposed. By comparing the area ratio of the hyper-entropy expected curve with the area and / or area As a measure of relevance, to solve the interval constraints of cloud model and the problem of semi-cloud measurement.The cloud entropy (3σ) constraint is used to increase the aggregation of cloud droplets and reduce the computational overhead.The hyper entropy is included in the calculation, and the cloud droplet thickness This method overcomes the problem of high computational complexity and unstable results in the distance metric method and the digital feature transform method for random cloud drops and satisfies the actual computational requirements of the three types of constraints.Experiments show that this method Can objectively and effectively measure the relevance of the cloud model, and has been verified in the cloud model-based system evaluation task.