基于贝叶斯网络的击剑训练优化

在击剑训练中,基于贝叶斯网络建立正向推理和逆向推理模型,发现训练过程和生理指标的相互关系.结合经验知识和样本数据,对该模型中网络结构的构造和网络参数的赋值方法进行详细说明.以女子重剑队的数据进行实验,并与BP神经网络方法进行性能比较.实验表明,该模型能有效的为教练员提供决策支持.     

基于贝叶斯网的象山港网箱养殖水环境指标建模

贝叶斯网络具有强大的推理能力,能与先验知识和数据结合,进行定性和定量分析,提供了1条有效的处理预测问题的途径,本文首先介绍了贝叶斯网络基本理论及其特点,并讨论如何学习贝叶斯网络结构,然后由专家知识和给定数据,采用基于依赖分析的贝叶斯网络学习算法构造了海底网箱养殖水环境指标间的贝叶斯网结构模型.该模型能有效的表达网箱养殖环境各个指标之间的因果关系和影响程度,实验结果表明,试验数据显示准确性是92.3%,kappa指数是0.882.以上证明该方法是有效可行的,表明贝叶斯网络是一种很有前途的预测评价方法.     

一个黄渤海海雾大气水平能见度算法

为了提高对黄渤海海雾天气海面大气水平能见度(以下简称"能见度")的数值预报能力,利用黄渤海23个沿岸和岛屿测站2013—2017年雾天的地面观测数据,构建了基于湿度信息的能见度算法(A-F算法),并将之应用于黄渤海海雾的能见度数值预报。结果表明,与常用的根据模式预报的云水含量诊断能见度的SW算法(Stoelinga and Warner,1999)相比,A-F算法表现更优,尤其可以诊断出被SW算法漏报的能见度为1～3 km的轻雾,说明A-F算法对黄渤海海雾天气能见度的数值预报具有一定应用价值。若将来加入浮标与船舶观测数据,可以进一步改进A-F算法能见度公式的具体形式;依据本文构建A-F算法的思路,可以发展适合其他海域的海雾天气能见度诊断公式。     

影响夏季北极航道航行的近地面气象要素时空变化特征分析

本文利用ERA5再分析数据和我国北极科学考察期间获取的走航气象观测数据,分析了夏季影响船舶通航北极航道的关键近地面气象要素的时空变化特征。结果表明,7–8月的天气条件最适宜船舶在北极航道航行,9月低温、大风和大浪天气显著增多,对船舶航行影响较大,10月的天气更加恶劣,对船舶航行的挑战更大。低温天气主要出现在各航道的中段,大风和大浪天气集中在航道两端的海域。除北极中心区和10月的挪威海和巴伦支海以外,其余时间的海域出现大风和大浪天气的概率以增加趋势为主,但具有较大的年际变化。根据现有北极航道气象观测数据分析发现,东北航道能见度最差,西北航道能见度最好,中央航道居中。     

基于多态模糊贝叶斯网络的半潜平台安装风险分析

针对半潜式平台安装作业风险评估中存在的风险多态性和模糊性问题,构建了一种多态模糊贝叶斯网络风险分析模型。根据行业规范推荐标准定义语言性评价模糊集,描述根节点的事故状态发生概率,克服了传统方法中确定性概率难以获取的困难。利用相似性聚合法结合置信度指标融合专家意见,引入改进的去模糊化转换方法,提高了专家知识经验转化为定量数据的合理性和可靠性。基于贝叶斯网络的双向推理和敏感性分析技术,实现了工程作业全过程的风险评估。通过对陵水17-2项目半潜平台整体吊装过程进行风险分析,验证该模型的合理性与有效性,为半潜平台安装作业风险管理与防控策略制定提供指导。     

基于加权贝叶斯网络的海洋灾害评估与管理

在全球气候变化背景下,海洋灾害的群发性、难以预见性和灾害链效应日显突出,带来的损失逐年上升,开展海洋灾害评估对于海洋经济建设、资源开发和工程建设具有重要的现实意义。文章首先基于风险理论剖析了海洋灾害风险的不确定性特征,构建了灾害评估指标体系;然后基于贝叶斯网络模型,提出了处理不确定性灾害评估的风险贝叶斯网络,进而基于主客观定权,构建了加权贝叶斯网络评估模型;最后对我国沿海地区海洋灾害开展评估研究。实验表明,该评估模型实现了海洋灾害的风险评估,具有一定的科学性和可行性。     

基于加权贝叶斯网络的海洋灾害风险评估

全球气候变化背景下,海洋灾害的群发性、难以预见性和灾害链效应日显突出,造成的损失逐年上升,开展海洋灾害的风险评估工作至关重要。针对海洋灾害评估中的不确定问题,本文首先基于风险理论剖析了海洋灾害风险的不确定性特征,构建了灾害评估指标体系;然后基于贝叶斯网络模型,提出针对不确定性灾害评估的风险贝叶斯网络,进而基于主客观定权,构建了加权贝叶斯网络评估模型;最后对我国沿海地区海洋灾害开展评估研究。实验表明,该评估模型有效实现海洋灾害的风险评估,具有实际可操作性。     

海域使用时空数据动态管理技术研究

运用面向对象方法的方法,对Geodatabase数据模型进行扩展,研究了海域使用时空数据的结构化存储技术,实现了历史数据的回溯和海域使用数据的变化跟踪等功能.结合厦门市海域使用动态管理的需求进行了应用研究,结果表明, Geodatabase 扩展数据模型研究成果很好地解决了宗海的历史回溯、变化跟踪、时态关系计算等问题.海域动态管理技术研究对海域管理具有重要意义.     

青岛市近海海域PM_(2.5)浓度及其与海面风和能见度的关系

文章基于小麦岛海洋环境监测站的相关数据,对2016年1-10月青岛市近海海域PM_(2.5)浓度的月平均变化特征、日平均变化特征和日变化特征进行研究,研究结果表明:PM_(2.5)浓度在20μg/m~3和30μg/m~3区间的发生天数最多,日均浓度最高值为277.7μg/m~3,重度污染共出现7d;PM_(2.5)浓度与近海海域气温和边界层高度有关,清晨偏高,午后偏低;PM_(2.5)浓度冬季普遍比夏季高,而夏季日较差比冬季大。结合对海面风的风速和风向以及海面能见度的观测,对PM_(2.5)浓度与二者的关系进行研究,研究结果表明:风速越大,PM_(2.5)浓度越小;PM_(2.5)浓度高值出现在西南风到西北风和东北风,谷值出现在东风到南风和北风,这与地形和海、陆PM_(2.5)浓度差异密切相关;PM_(2.5)浓度越高,海面能见度越低,且可利用指数曲线估计海面能见度。     

Jason-2有效波高产品在中国海域的准确性评估

Jason-2卫星高度计的有效波高(Hs)产品在海洋学领域得到了广泛应用。为了综合评估Jason-2有效波高产品在中国多个海域和不同海况下的准确性,为其在中国海域的应用研究提供参考,本文基于国家海洋局20个水文气象浮标从2008年至2014年的长期观测数据,对Jason-2有效波高产品在中国三个海区(渤黄海、东海、南海)和不同海况(观测有效波高Hs最大为6.2m)下的准确性进行综合分析,并研究其准确性与离岸距离的关系。结果表明Jason-2有效波高产品:(1)在中国海域具有较好的准确性,均方根误差RMSE=0.445m;(2)在南海的准确性最高,而在渤黄海的准确性较差;(3)在2m≤Hs≤5m的海况条件下准确性较好;(4)在越远离陆地的海域准确性越高。     

Prediction of visibility in the Arctic based on dynamic Bayesian network analysis

With the accelerated warming of the world, the safety and use of Arctic passages is receiving more attention.Predicting visibility in the Arctic has been a hot topic in recent years because of navigation risks and opening of ice-free northern passages. Numerical weather prediction and statistical prediction are two methods for predicting visibility. As microphysical parameterization schemes for visibility are so sophisticated, visibility prediction using numerical weather prediction models inclu...     

基于贝叶斯网络的海洋生态环境状况评价——以山东省为例

在海洋开发利用活动日益增多,海洋环境保护压力日益增大的背景下,进行海洋生态环境状况科学评价方法的研究可以为开展海洋生态环境监测评估与海洋生态环境保护政策制定提供科学依据,促进改善海洋生态环境状况。本文首先基于贝叶斯网络基本理论,构建了海洋生态环境状况贝叶斯网络评价模型;再以山东省为例,充分融合专家经验知识与客观数据,从定性与定量两个角度评价了2013—2019年山东省近海生态环境状况,分析了各评价指标间关系及其对近海生态环境状况的影响程度。评价结果表明2013—2019年山东省近海生态环境状况保持一般状态,海洋环境质量与海洋生态灾害类指标对评价结果影响最大,山东省海洋环境保护工作初见成效。研究结果表明该评价模型能有效评价海洋生态环境状况,分析结果对于环境保护政策制定具有一定的借鉴意义。     

Otolith reading and multi-model inference for improved estimation of age and growth in the gilthead seabream Sparus aurata (L.)

Accurate knowledge of fish age and growth is crucial for species conservation and management of exploited marine stocks. In exploited species, age estimation based on otolith reading is routinely used for building growth curves that are used to implement fishery management models. However, the universal fit of the von Bertalanffy growth function (VBGF) on data from commercial landings can lead to uncertainty in growth parameter inference, preventing accurate comparison of growth-based history traits between fish populations. In the present paper, we used a comprehensive annual sample of wild gilthead seabream (Sparus aurata L.) in the Gulf of Lions (France, NW Mediterranean) to test a methodology improving growth modelling for exploited fish populations. After validating the timing for otolith annual increment formation for all life stages, a comprehensive set of growth models (including VBGF) were fitted to the obtained age–length data, used as a whole or sub-divided between group 0 individuals and those coming from commercial landings (ages 1–6). Comparisons in growth model accuracy based on Akaike Information Criterion allowed assessment of the best model for each dataset and, when no model correctly fitted the data, a multi-model inference (MMI) based on model averaging was carried out. The results provided evidence that growth parameters inferred with VBGF must be used with high caution. Hence, VBGF turned to be among the less accurate for growth prediction irrespective of the dataset and its fit to the whole population, the juvenile or the adult datasets provided different growth parameters. The best models for growth prediction were the Tanaka model, for group 0 juveniles, and the MMI, for the older fish, confirming that growth differs substantially between juveniles and adults. All asymptotic models failed to correctly describe the growth of adult S. aurata, probably because of the poor representation of old individuals in the dataset. Multi-model inference associated with separate analysis of juveniles and adult fish is then advised to obtain objective estimations of growth parameters when sampling cannot be corrected towards older fish.     

Human–environment interaction in the Baltic Sea

A model based on a Bayesian Belief Network (BBN) has been constructed for the Baltic Sea with the aim of investigating future scenarios of human activities in the region and informing environmental management strategies, such as those developed under a Science and Policy Integration for Coastal Zone Assessment Systems Approach Framework application. This paper describes necessary refinements to take into account historical influences on this relatively enclosed system. BBNs are static models and therefore do not incorporate feedback loops, whereas natural systems clearly display feedback mechanisms. This paper describes the implementation of one step feedback loops into a BBN model in an attempt to partly remove this constraint. Feedback loops within this stochastic model were shown to improve its accuracy. The drivers, both natural and anthropogenic, having greatest impact on the environment are identified. These refinements were made to improve its accuracy in modelling the system and gives insights into the functioning of that system.     

建立海洋平台管节点PSN曲线的贝叶斯方法

针对管节点疲劳试验的小样本特点,探讨了建立ＰＳＮ 曲线的贝叶斯方法。在贝叶斯方法中,ＰＳＮ曲线的统计参数作为随机变量处理,首先根据贝叶斯定理求出参数向量的后验概率密度,然后建立ＰＳＮ曲线的贝叶斯方程,最后再编程计算。算例表明,与传统的ＰＳＮ 曲线相比,贝叶斯ＰＳＮ 曲线更加安全可靠     

Evolving a Bayesian network model with information flow for time series interpolation of multiple ocean variables

Based on Bayesian network (BN) and information flow (IF), a new machine learning-based model named IFBN is put forward to interpolate missing time series of multiple ocean variables. An improved BN structural learning algorithm with IF is designed to mine causal relationships among ocean variables to build network structure. Nondirectional inference mechanism of BN is applied to achieve the synchronous interpolation of multiple missing time series. With the IFBN, all ocean variables are placed in a causal network visually, making full use of information about related variables to fill missing data. More importantly, the synchronous interpolation of multiple variables can avoid model retraining when interpolative objects change. Interpolation experiments show that IFBN has even better interpolation accuracy, effectiveness and stability than existing methods.     

Markov chain Monte Carlo (MCMC) sampling methods to determine optimal models,model resolution and model choice for Earth Science problems

We present an overview of Markov chain Monte Carlo, a sampling method for model inference and uncertainty quantification. We focus on the Bayesian approach to MCMC, which allows us to estimate the posterior distribution of model parameters, without needing to know the normalising constant in Bayes' theorem. Given an estimate of the posterior, we can then determine representative models (such as the expected model, and the maximum posterior probability model), the probability distributions for individual parameters, and the uncertainty about the predictions from these models. We also consider variable dimensional problems in which the number of model parameters is unknown and needs to be inferred. Such problems can be addressed with reversible jump (RJ) MCMC. This leads us to model choice, where we may want to discriminate between models or theories of differing complexity. For problems where the models are hierarchical (e.g. similar structure but with a different number of parameters), the Bayesian approach naturally selects the simpler models. More complex problems require an estimate of the normalising constant in Bayes' theorem (also known as the evidence) and this is difficult to do reliably for high dimensional problems. We illustrate the applications of RJMCMC with 3 examples from our earlier working involving modelling distributions of geochronological age data, inference of sea-level and sediment supply histories from 2D stratigraphic cross-sections, and identification of spatially discontinuous thermal histories from a suite of apatite fission track samples distributed in 3D.