用信息扩散模型改进台风风险估计

频繁且随机发生的台风,其风险是概率风险,具有统计学的意义.用历史台风的损失记录作为样本来量化风险时,台风风险就是台风损失样本的平均值,它必等于由台风发生概率和易损性偶合估计出的损失期望值.当统计资料较少时,由于信息扩散模型能提高小样本的工作效率,消除用简单统计方法估计风险时出现的平缓和跳空现象,较好地表现统计规律,从而能改进台风风险的估计精度.按年分段形成的多个样本函数可用于计算物理意义明确的台风年度风险值,并可用信息扩散模型对其加以改进.蒙特卡罗法并不具有改进风险评估精度的功能.本文以浙江省部分历史台风资料为样本,演示了台风风险的估计.信息扩散模型的改进程度如何,将另文用计算机仿真的方法来加以检验.

Improving estimation of typhoon risk by using information diffusion model

The risk of typhoon occurring frequently and randomly is probability risk that has a statistical significance. When we quantify a typhoon risk by using loss records of historical typhoons as a sample, the typhoon risk is the average of the sample. The average is equal to the expected value of the loss that is estimated by coupling the probability of occurrence of typhoon and the vulnerability. When the statistical data is not much, the information diffusion model can improve the work efficiency of the sample, eliminate the flat and gap phenomenon in estimating risk with a simple statistical method, and more better show the statistical regulation, so that the model can improve the accuracy of estimating typhoon risk. A number of sample functions formed by yearly segmentation can be used to calculate the typhoon risk per year of a clear physical meaning. The information diffusion model also can improve the estimation of the risk per year. The Monte Carlo method cannot improve the accuracy of risk assessment. In this paper, some historical typhoon data in Zhejiang province are used as samples to demonstrate the estimation of typhoon risk. Using the computer simulation method in a subsequent article, we will show how degree the information diffusion model can improve an estimation of typhoon risk.