北京市PM2.5-O3复合污染数值响应解析

为分析PM_2.5-O₃复合污染数值响应关系,基于2015～2020年北京市空气质量数据、气象资料和新冠疫情数据,分析PM_2.5-O₃复合污染事件在多尺度下的变化趋势.同时提出一种复合污染指数,在广义相加模型下分析数值响应趋势,并进一步引入分布滞后非线性模型,分别解析复合污染指数、复合污染事件和影响因素间的滞后响应关系.结果表明,北京市PM_2.5-O₃复合污染事件逐年减少,具有明显的季节效应、星期效应和节假日效应.复合污染指数与降雨量无明显相关性,与O₃和空气温度呈线性正相关,与其余解释变量均为非线性相关.同时大气污染物和气象条件对复合污染指数有明显滞后效应,滞后影响主要集中在1～3 d.高值的PM_2.5、 PM₁₀、 O₃、 SO₂和空气温度明显增加复合污染风险,中值段的CO(1～6 mg·m^-3)、 NO₂     

基于Kinect的地铁乘客不安全行为识别方法与实验*

为解决地铁视频监控技术对乘客不安全行为只记录不识别且较少考虑识别精确度的问题,提出1种基于Kinect传感器的高效识别方法。以Pelvis为向量起点和动作活动高频关节为终点构建识别特征向量;运用余弦定理获得标准动作与测试动作关节的最大角度差序列;以最大角度差为动作特征量建立相似度计算模型,运用动态时间规整算法（DTW）将初始结果转换为动作相似度。以相连关节法为对照组开展对比实验,结果表明:前者在抽烟、挥拳、挥手呼救等行为识别的准确度分别为91.7%,86.9%,89.2%,平均比对照组高4%以上,显著提高了地铁乘客不安全行为的识别率,可为地铁智能安全管控提供理论与技术依据。     

遗传算法在多目标土地利用规划中的应用

在建立和求解多目标土地利用规划模型过程中，采用遗传算法并结合多目标模糊优选理论处理多目标土地利用结构优化问题。其详细步骤为：在分析区域特征基础上构建时间序列；通过现状分析设置土地利用类型决策变量、确定优化目标和约束条件。构建土地利用结构多目标优化模型；用灰色GM（1，1）预测模型对土地利用结构优化多目标模型参数进行预测；用遗传算法多目标优化方法对土地利用结构优化模型求解得pareto解集；最后运用多目标模糊动态优选方法对规划期内各时间段的pareto解集进行模糊动态优选，以最大优属度判定原则确定出最优决策序列。自贡实例表明，该方法有助于解决多目标协调问题，通过遗传算法避免了人为偏好权重的主观性，并通过提供多方案选择性，从而使规划更具灵活性，同时又便于充分吸纳公众参与，减少规划中的不确定性，增加规划的科学性、现实性和可操作性。     

风化煤腐植酸增效尿素红外光谱分析

采用碱性混合活化剂处理风化煤,研制腐植酸添加剂,按不同比例添加到尿素熔融液中,采用模拟喷浆造粒的方法,制备3类增效尿素。通过风化煤与添加剂、尿素与增效尿素红外光谱特征对比发现：风化煤经过活化后,碳单键数量降低,碳链缩短,活性腐植酸HA类提取物出现活性官能团,FR类除单键数量减少和碳链缩短外,生成部分胺、酰胺结构;增效剂与尿素在反应过程中,出现了三键和累积双键破坏、稳定性复合物形成、碳链缩短、双键结构增加等变化,但不同类型增效剂对产品红外光谱特征影响存在差异。     

基于MCOA布置矿井瓦斯传感器

为了预防矿井瓦斯事故,确定增设瓦斯传感器的数量和位置。通过建立监测有效 等级来确定增设瓦斯传感器的数量,然后通过分析多个猴群同步并行搜索、简易猴群算 法的初步搜索和基本和声算法的二次搜索等方法,设计出猴群优化算法的计算流程,再 根据瓦斯传感器布置的目标函数确定增设瓦斯传感器位置。以大隆矿为例进行了现场工 业试验,结果表明,增设瓦斯传感器的数量与监测有效等级之间成正比例关系,但并非 随着监测有效等级的提高而持续增加。     

改进的GA-BP神经网络在矿井突水水源判别中的应用

矿井突水水源的判别是制定防治水措施的重要环节。通过对某矿含水层水化学特 性的相关性分析,将PCA算法、K折交叉验证算法嵌入GA-BP神经网络,提出了一种新的 GA-BP神经网络,将其应用于实例分析中,并与传统的方法进行比较。结果表明:针对 水化学特性相近的含水层,PCA算法能够排除样本中的冗余信息,降低样本指标维度, 简化BP神经网络结构;K折交叉验证算法能够提高GA算法对BP神经网络权值的寻优质量 ,使GA算法的进化方向更具合理性;二者的引入大大优化了传统GA-BP神经网络性能, 其判别精度更高、适用性更强、结果更可靠,在矿井突水水源判别方面具有很好的应用 前景。     

Land-use planning via enhanced multi-objective evolutionary algorithms: optimizing the land value of major Greenfield initiatives

The successful implementation of major development initiatives relies on the sound allocation of land uses against critical design criteria and constraints. The discovery of optimum development plans introduces severe complexities in formulating and solving the underlying multi-objective optimization problem. Moreover, in the presence of conflicting planning criteria decision-makers should be provided with a set of alterative-yet-optimum solutions that uniformly cover the spectrum of feasible maps. The introduction of sophisticated optimization algorithms addresses this challenge by pursuing a complete approximation of the Pareto front containing all prominent spatial allocations. This study demonstrates the effectiveness of a new evolutionary algorithm (UDT-MOEA) against the results of an established multi-objective genetic algorithm (NSGA-II) when applied on a major greenfield initiative against the optimum location(s), size and shape of three new land uses. Each algorithm performs best in different areas of the feasible objective space, providing planning alternatives with distinct characteristics.     

Positive relationships between phenol oxidase activity and extractable phenolics in estuarine soils

Decomposition of recalcitrant materials such as phenolics is known to play a pivotal role in organic matter decomposition and nutrient cycling in estuaries. The specific goals of this study were to determine temporal and spatial variations of phenol oxidase and phenolics in estuarine soils, and to elucidate controlling factors for phenol oxidase activity. To achieve these goals, phenol oxidase activity and phenolic content were measured in soils developed along the side of an estuary in the Han River, Korea. Soil samples were collected in three locations with different vegetation: mud flats, Zizania-dominated soils, and Salix-dominated soils. Monthly measurements were also made in a Zizania-dominated site over a year period. Phenol oxidase activity varied between 0.00 and 0.28 diqc min^?1 g^?1, whilst phenolic content ranged from 0.0–10.5 μg g^?1. A correlation analysis revealed that phenol oxidase activity exhibited positive correlations with phenolic content in both seasonal and spatial data. The same relationship was found when the data were analysed separately for each site. Unlike peatlands or upland forest soils where negative correlations were often found between phenol oxidase activity and phenolics, substrate induction appears to account for the positive correlation in the present study.     

Carboxymethyl‐β‐cyclodextrin influence on the solubility and toxicity of substituted indole compounds

The solubilization of eight selected substituted indole compounds (SICs) by carboxymethyl‐β‐cyclodextrin (CMCD) in water is reported. The results show that their solubility was well improved in CMCD aqueous solution, and the stoichiometry of inclusion complex of the studied SICs with CMCD was 1: 1. The acute toxicity to Photobacterium phosphoreum of the investigated SICs in CMCD aqueous solutions was also investigated and was compared with that in water. The results show that the toxicity of five studied SICs decreased while the toxicity of the rest three SICs increased in CMCD aqueous solutions, but neither the increase nor the decrease was significant.     

Long-term prediction model of rockburst in underground openings using heuristic algorithms and support vector machines

Rockburst possibility prediction is an important activity in many underground openings design and construction as well as mining production. Due to the complex features of rockburst hazard assessment systems, such as multivariables, strong coupling and strong interference, this study employs support vector machines (SVMs) for the determination of classification of long-term rockburst for underground openings. SVMs is firmly based on the theory of statistical learning algorithms, uses classification technique by introducing radial basis function (RBF) kernel function. The inputs of models are buried depth H, rocks’ maximum tangential stress σ_θ, rocks’ uniaxial compressive strength σ_c, rocks’ uniaxial tensile strength σ_t, stress coefficient σ_θ/σ_c, rock brittleness coefficient σ_c/σ_t and elastic energy index W_et. In order to improve predictive accuracy and generalization ability, the heuristic algorithms of genetic algorithm (GA) and particle swarm optimization algorithm (PSO) are adopted to automatically determine the optimal hyper-parameters for SVMs. The performance of hybrid models (GA + SVMs = GA-SVMs) and (PSO + SVMs = PSO-SVMs) have been compared with the grid search method of support vector machines (GSM-SVMs) model and the experimental values. It also gives variance of predicted data. A rockburst dataset, which consists of 132 samples, was employed to evaluate the current method for predicting rockburst grade, and the good results of overall success rate were obtained. The results indicated that the heuristic algorithms of GA and PSO can speed up SVMs parameter optimization search, the proposed method is robust model and might hold a high potential to become a useful tool in rockburst prediction research.