基于粗糙集与灰色SVM的中国CO2排放量预测

从我国CO2排放量的不确定性、不完整性、小样本等特征出发,以灰色系统(GM)模型和支持向量机(SVM)模型为基础,建立基于粗糙集的组合预测模型.利用该模型以我国1990~2011年CO2排放量的数据以及同期的人口数量、GDP和能源消耗总量数据为基础对我国同期CO2排放量进行预测来验证其有效性,最后对我国2012~2017的CO2排放量进行预测.结果表明,灰色系统理论与支持向量机模型仅能够反应我国CO2排放的长期变化趋势,在预测精度上存在一定缺陷而基于粗糙集与灰色SVM的组合预测模型在预测精度上明显优于以上两种方法,能够对我国未来CO2排放量进行准确有效的预测分析.     

Lessons learned for process safety management in China

A number of chemical accidents have occurred in China over the past two decades with significant impact on humans and the environment. It is expected that lessons will have been learned from these accidents that will help industries to reduce the risk that catastrophic chemical accidents occur in future. In fact, to some extent there is evidence that lessons have been learned, to the extent that the Chinese government has substantially strengthened legislation and regulatory standards. Nonetheless, there remains a concern that much more still needs to be done to reduce chemical accidents risks in China. Important progress in this area requires not only government support but a commitment across all hazardous industries to learn from past accidents that may in many cases require establishment or considerable improvement of their safety management systems. To assist small and medium-sized enterprises (SMEs), in this effort, results of an analysis of common causes of the chemical accidents reported in the Major Accident Information (MAI) website of Chinese State Administration of Work Safety (SAWS) are presented in this paper In particular, inadequate process hazard analysis (PHA), training and emergency response planning (ERP) were identified as the top three process safety management (PSM) elements that contribute to most of the SMEs accidents in China. Seven recommendations are proposed in order to improve the effectiveness of lesson learning for government agencies and SMEs.     

矿山振动机械本质安全技术研究及应用

目前矿山开采领域所使用的振动机械(包括振动磨、振动筛等)的振动负荷都无一例外地传递给弹簧、机架、基础,对机器相关零部件及周围会造成不同程度的安全影响,安装要求也较高。从设备的安全性角度来说,无法达到本质安全要求。按照传统振动机械的设计理论,振动负荷向外传递无法避免,无法实现本质安全。提出振动机械设计理论,可以实现本质安全化设计。实际应用结果表明,外传的振动负荷降低95%以上,且可达到节能50%以上。     

A density projection approach for non-trivial information dynamics: Adaptive management of stochastic natural resources

We demonstrate a density projection approximation method for solving resource management problems with imperfect state information. The method expands the set of partially-observed Markov decision process (POMDP) problems that can be solved with standard dynamic programming tools by addressing dimensionality problems in the decision maker's belief state. Density projection is suitable for uncertainty over both physical states (e.g. resource stock) and process structure (e.g. biophysical parameters). We apply the method to an adaptive management problem under structural uncertainty in which a fishery manager's harvest policy affects both the stock of fish and the belief state about the process governing reproduction. We solve for the optimal endogenous learning policy—the active adaptive management approach—and compare it to passive learning and non-learning strategies. We demonstrate how learning improves efficiency but typically follows a period of costly short-run investment.     

基于QSPR方法的脂肪醇化合物闪点预测

为提高脂肪醇化合物闪点预测精度,提出基于定量结构-性质关系(QSPR)原理的脂肪醇化合物闪点预测方法。应用Dragon软件计算出91种脂肪醇的分子描述符,利用遗传函数算法(GFA)从1 481个描述符中筛选出3个与脂肪醇闪点关系最密切的分子描述符。分别用多元线性回归(MLR)方法和支持向量机(SVM)方法进行建模,并采用内部验证和外部检验的方式对模型的拟合度、预测性等性能进行验证。结果表明:预测集的MLR方法和SVM方法的平均绝对误差(AAE)分别为2.870 K和2.706 K;均方根误差(RMSE)为3.451 K和3.371 K。SVM模型在精度上略优于MLR模型,而MLR模型更为简单和方便。     

基于高斯过程机器学习算法的矿柱稳定性分析

矿柱是地下矿山支撑顶板围岩、维持采场稳定的关键结构要素。为迅速准确地判别矿柱稳定性,选取矿柱宽度、矿柱高度、矿柱宽高比、矿岩单轴抗压强度和矿柱承受载荷作为影响指标,利用高斯过程机器学习算法建立矿柱状态与其主要影响因素之间的非线性映射关系,进而提出一种基于高斯过程二元分类(GPC)的矿柱状态识别模型。结合工程实例,以40组样本数据进行训练,以10组样本数据对该模型进行检验,并与ANN和SVM进行对比。结果表明,矿柱状态识别的高斯过程模型是科学可行的,该模型具有参数自适应化获取、分类精度高、计算复杂度低等优点,还可对矿柱状态判别结果的不确定性或可信度进行定量化评价。     

模糊控制理论在湿法脱硫真空皮带脱水机滤布纠偏控制中应用

随着国家节能减排工作的不断深入,燃煤电站脱硫装置的投运容量已占全国燃煤机组的一半左右。靖远第二发电公司4×300MW机组脱硫装置自投运以来多次出现真空皮带脱水机滤布跑偏的现象,造成真空皮带过滤机停车及滤布撕裂等严重后果,影响脱硫装置的正常投运。介绍了通过引进先进的模糊控制理论应用于滤布纠偏控制,并对控制硬件进行相应的改造,成功地解决了靖远第二发电公司湿法脱硫真空皮带脱水机滤布纠偏控制中出现的问题。为燃煤电站湿法脱硫真空皮带脱水机滤布及其他行业相似工艺过程的纠偏控制提供参考。     

烧结机湿法烟气脱硫吸收塔本体设计与应用

依据烧结烟气的特性,选用旋流板塔为脱硫吸收设备并对其结构进行设计。通过工程实例说明设计的塔型具有脱硫效率高、液气比低、运行费用低等优点,是一种值得应用于烧结机烟气脱硫的装置。     

基于LSSVM的石灰石浆液制取电耗优化

提出了基于最小二乘支持向量机(LSSVM)构造以球磨机制浆速度作为目标优化模型,并采用遗传算法实现了给料量、给水量、漩流器入口压力等运行参数的寻优,获得了优化制浆速度的调整方式。以此方式的相应参数作为当前负荷下的基准值,能很好地解决制浆工况下运行参数基准的确定问题。通过现场试验证明:按此基准值运行,可以提高制浆速度,降低电耗。     

Sensori-motor learning and its relevance for task specialization in bumble bees

Individual bees often restrict their visits to only a few species out of the multitude of available plants. This flower constancy is likely caused by limitations of memory for motor patterns, sensory stimuli, or reward levels. Here we test the implications of sensori-motor learning and memory for flower constancy. Artificial “flowers” with two distinct “morphologies” were used, so that in each flower type, a different motor pattern was needed to reach the nectar. As in natural flowers, these morphological types were associated with sensory signals (blue and yellow color stimuli). Bees which learned only a single task were more efficient in several ways than those which had learned two: they made fewer errors, had shorter flower handling times, took shorter times to correct errors, and transitions between flowers were initially more rapid. For bees which had learned two tasks, performance depended strongly on the training schedule: if each task was learned with blocked trials, the memory for the second appeared to interfere with that for the first. Interference affected only the association between flower signal and motor pattern, not the motor pattern itself. This was not the case if bees were trained for both tasks with alternating trials. In that case, bees rapidly learned both tasks, albeit with worse saturation levels than bees which had learned only one. Bees transferred the experience gained on one task to a second task: their initial performance on the second task was better than their initial performance on the first. On the other hand, performance on the second task in the saturation level (in which bees no longer improve their efficiency) was worse than on the first task (negative transfer). In the saturation phase, performance did not directly depend on switch frequency, but on whether the bee had one or two options in memory. Thus, while bees would become proficient at two tasks more quickly if their acquisition phase included switches, such switches had no measurable effect in the saturation phase. The implications of these findings for foraging are discussed using modern learning theory. Received: 4 April 1997 / Accepted after revision: 8 August 1997