边坡变形时序非线性判定及混沌预测研究

以探讨边坡变形性质及混沌预测可行性为目的,基于混沌理论利用相空间重构技术对其变形时间序列进行混沌特征判定,试验显示变形系统具有混沌特性,可用混沌相关理论进行研究;基于混沌相空间重构技术,笔者构建了多种混沌预测模型进行混沌预计研究,分析各类模型的工程实际应用效果;针对单次监测时序预测精度较低的问题,提出累加时序预测方案,训练结果显示,短期预测精度变形累计值基本控制在5%以内,高程值预测相对误差均低于1%,预测精度较高,可以用于工程实际。     

基于人工神经网络理论的建筑物火灾安全评价研究

依据建筑物火灾危险性的影响因素,应用人工神经网络理论及系统安全方法,建立了建筑物火灾危险性的评价指标体系,该方法摆脱了评价过程中的随机性和参评人员主观上的不确定性及其认识上的模糊性等缺点,大大提高了准确性。为了验证评价模型的准确性,将该理论应用到某高校图书馆火灾危险性评价中,快速、准确地得到了安全评价结果,取得了满意效果,为建筑物防火设计以及安全管理提供了可行的依据。     

A process integration approach to industrial water conservation: a case study for a Chinese steel plant

A systematic approach to optimizing water network has traditionally been utilized to exam and plan water conservation in industrial processes. In the present case study, water-pinch technology was used to analyze and optimize the water network of a steel plant near China's Zhangjiakou city. A system design was developed and a limiting constraint (Cl(-) concentration) was identified based on investigations of water quality then the minimum freshwater and wastewater targets were determined without considering water losses. The analysis was then extended by calculating the additional input of freshwater required to balance the actual water losses. A nearest-neighbor algorithm (NNA) was used to distribute the freshwater and recycled water among each of the plant's operations. The results showed that with some reconstruction of the water network, the flow rates of freshwater and wastewater could be decreased by 57.5% and 81.9%, respectively.     

Modeling the Effects of Potential Salinity Shifts on the Recovery of Striped Bass in the Savannah River Estuary,Georgia–South Carolina,United States

Increased salinity in spawning and nursery grounds in the Savannah River estuary was cited as the primary cause of a 97% decrease in adult striped bass (Morone saxatilis) and a concomitant 96% decrease in striped bass egg production. Restoration efforts focused on environmental remediation and stock enhancement have resulted in restored salinity patterns and increased egg and adult abundances. However, future water needs or harbor development may preclude further recovery by reducing freshwater inflow or increasing salinity intrusion. To assess the effect of potential changes in the salinity regime, we developed models relating discharge, tidal phase, and salinity to striped bass egg and early larval survival and re-cast these in a quantitative Bayesian belief network. The model indicated that a small upstream shift (≤1.67 km) in the salinity regime would have the least impact on striped bass early life history survival, whereas shifts >1.67 km would have progressively larger impacts, with a 8.33-km shift potentially reducing our estimated survival probability by >28%. Such an impact could have cumulative and long-term detrimental effects on the recovery of the Savannah River striped bass population. The available salinity data were collected during average and low flows, so our model represents some typical and some extreme conditions during a striped bass spawning season. Our model is a relatively simplistic, “first-order” attempt at evaluating potential effects of changes in the Savannah River estuarine salinity regime and points to areas of concern and potential future research.     

浅谈建筑施工噪声的网络化管理

通过《南宁市环境信息综合管理系统》中建筑施工噪声管理子系统的重点研究和建设,对建设施工项目审批、中午夜间施工许可登记及施工噪声投诉实时监控,加强监督,改善了管理环节,提高了管理效能,取得了良好的效果。     

基于BP模型对城市交通噪声的数据处理和预测

城市交通噪声的预测和评价技术是城市交通可持续发展的重要研究内容,直接为城市交通规划中环境容量分析和环境影响评价服务。本文通过实测的大量数据,运用神经网络中的BP模型及其算法建立车辆数、道路宽度和交通噪声之间的关系,对城市道路交通噪声数据进行处理和预测。     

江苏数字地震台网地震定位能力分析

依据江苏数字地震台网的几何布局 ,对区域上设定的不同震级和震源位置的可能事件 ,作了以下两点考虑。 (1)根据不同震级地震的监测范围 ,确定了相应的定位子台组合。 (2 )对同一地震事件 ,考虑到震相分析P震相和S震相测量精度上的差异 ,引入了不同的标准误差σP 和σS。在此基础上 ,通过模拟地震波理论走时和对走时矩阵的奇异值分解方法 ,计算了估计震源参数标准误差的参量空间协方差矩阵值 ,给出了标准误差理论值等值线图 ,用此方法讨论分析了江苏数字地震台网的地震定位能力。     

江苏数字地震台网台址噪声谱分析

江苏数字地震台网已建设改造了 8个台站。对 8个台站记录到的背景噪声进行定量分析 ,测定背景噪声的地动速度均方根值 (RMS)、有效测量动态范围 (D有效)和噪声功率谱。结果表明 :受台基和海潮等观测环境的影响 ,大部分台站的地动速度RMS值高于 2 × 1 0 -8m/s,降低D有效 值。背景噪声功率谱峰值频率为 2 5～ 3 0Hz。这些定量分析为了解数字地震台网的实际监测水平提供了可靠依据     

基于3种时间序列模型的九龙江河流库区藻华预测

湖库富营养化和有害藻华是全球性生态环境问题,藻华预测与早期预警是保障湖库水源地供水安全的关键技术.如何基于高频水生态在线监测数据进行藻华的实时动态预测成为水生态管理领域的重大需求.本研究以福建省九龙江江东库区(水源地)为例,利用3年连续观测的逐时平均总叶绿素a浓度数据,对比研究了SARIMA、Prophet和LSTM(长短期记忆神经网络)3种时间序列模型在藻华(日平均叶绿素a大于15μg·L^-1)预测方面的效果.结果表明：(1)时间序列模型要求参数少,灵活性强,能清晰反映水质特征和未来变化趋势,可弥补传统藻类监测预警方法的局限性;(2)基于深度学习框架的LSTM模型,具有独特的迭代优化算法,对藻类非线性变化特征的识别和预测能力较强,其总叶绿素a逐日预测和7日预测效果均显著优于SARIMA模型和Prophet模型;(3)输入数据长度会在一定程度上影响模型预测效果,最优的输入数据时间长度为7 d;输入数据频率对预测效果也有影响,在预测非藻华日时,小时数据的预测效果优于日频率数据;在预测藻华日时,两种频率数据无显著差异,但日频率数据能更准确识别藻华日特征.总结起来,基于...     

新疆环境监测网络的建设及发展对策

新疆环境监测网络建立三十多年来,已经初步形成了水环境、环境空气、声环境、生态环境和污染源等监测网络,较好地承担了各环境要素例行监测、突发性环境污染事故应急监测和环境科研等监测任务,为各级环保决策提供了有效的技术支持。然而,随着环境管理工作的不断深入,新疆环境监测网络所呈现的一些新老问题已成为不适应环境管理需求的制约因素,本文通过分析新疆环境监测网络中存在的问题,提出了发展对策。