Performance of CERES-Rice and CERES-Wheat models in rice–wheat systems: A review

Rice–wheat (RW) systems are critical to food security and livelihoods of rural and urban poor in south Asia and China, and to regional economies in southeast Australia. The sustainability of RW systems in south Asia is, however, threatened by yield stagnation or decline, and declining partial factor productivity, soil organic C and water availability. Crop models potentially offer a means to readily explore management options to increase yield, and to determine trade-off between yield, resource-use efficiency and environmental outcomes. This paper reviews the performance of CERES-Rice and CERES-Wheat in Asia and Australia in relation to their potential application towards increasing resource use efficiency and yield of RW systems.

The performance of the models was evaluated using simulated and observed data on anthesis and maturity dates, in-season LAI and growth, final grain yield and its components, and soil water and N balances from published studies across Asia and Australia, and then by computing the statistical parameters for the major characters. Over the four data sets examined for anthesis and six for maturity dates, CERES-Rice predicted those dates fairly well (normalised RMSE = 4–5%; D-index = 0.94–0.95), but over the 11 sets for grain and 4 for biomass yield, the predictions were more variable (normalised RMSE = 23% for both; D-index 0.90 and 0.76, for grain and biomass, respectively). Model performance was poorer under conditions of low N, water deficit, and low temperatures during the reproductive stages. Over the three data sets examined, CERES-Wheat predicted the anthesis and maturity dates quite well (normalised RMSE = 4–5%; D-index = 0.94–0.99), and over eight sets for grain and two sets for biomass yield the model predicted them also reasonably well (RMSE = 13–16%; D-index = 0.86–0.97). Only one study evaluated the DSSAT RW sequence model with fairly satisfactory predictions of rice and wheat yields over 20 years with adequate N, but not the long-term change in soil organic C and N. Predictions of in-season LAI and crop growth, and soil and water processes were quite limited to investigate the robustness of model processes.

Application of models to evaluate options to increase water and N use efficiency requires the ability to perform well at the margin where deficit stress begins. While both models generally perform satisfactorily under water and N non-limiting conditions, the little evidence available suggests that they do not perform well under resource-limiting situations. We recommend that the models’ key processes under the water and N limiting conditions be further evaluated urgently. The DSSAT sequence model also needs to be further evaluated against observations for a range of locations and management using data from long-term experiments in RW systems.     

齿轮三维模型的参数化设计技术

齿轮是汽车、拖拉机中常用的传动零件,在三维设计中,需要对渐开线方程进行转换,由转换后的方程生成齿形,其整个设计过程非常繁琐。本文详述了方程的转换过程及齿轮的参数化设计方法,可以通过输入不同的齿轮参数,即可获得相应的齿轮,从而大大节约设计时间。该方法还可应用于所有系列化的通用件和标准件三维设计中。     

水泵进水设计模型试验方法

模型试验是国际上普遍采用的检验和改进泵站进水设计的主要方法,而试验标准则是水泵进水设计模型试验的重要依据;目前我国尚未有关于泵站进水设计模型试验的试验标准。本文简要讨论了水泵进水设计的整体概念,并根据美国国家标准《水泵进水设计》(ANSI／HI9．8-1998),介绍了水泵进水设计模型试验的目的、主要内容、试验准则、测试方法、试验方案和试验结果评价等主要内容。     

基于PPC模型的农机选型与优序关系研究

将投影寻踪（PP）与实数编码的遗传算法（RAGA）相结合，应用到农业机械的选型与优序关系研究中，通过寻求各种机型评价指标的最佳投影方向，将高维数据转换成一维投影指标值，实现对机器样本的分类与排序，从而克服 以往二阶模糊综合评判法及灰色系统评价法中权重取值的人为因素，取得了良好的效果。     

管网中泵破坏瞬变流的数值模拟

建立了管网中泵破坏瞬变流的数值模拟，该模型考虑了泵破坏无量纲非线性边界条件，运用特征线法对方程进行了求解。经验证，证明模型是可靠的，可用于工程分析计算。     

灌区计算机监控系统与灌溉优化调度模型集成应用研究

大型灌区实施计算机监控是提高灌区管理水平的重要手段,灌区轮灌分组优化模型旨在求解干渠各出水口运行时调度的最优组合方案,而灌溉优化调度方案在实施过程中往往仍然采用的是计算机远程人工调度.为此研究了一种灌区轮灌分组优化模型,并得出了灌溉优化调度方案,同时采用OPC（即OLE for Process Control）技术,使灌溉优化调度方案数据与计算机控制系统有效地结合起来,提高了灌区灌溉管理自动化水平。     

大型复杂机械节点导入建模及有限元分析

针对一般有限元软件在大型复杂机构建模和划分网格方面的局限性,提出了节点导入有限元建模法。采用三维建模软件建立节点布局图,并提取源坐标数据,结合有限元分析软件建立有限元模型,同时完成网格划分,并成功地将此建模方法应用于大型振动筛的建模与分析。结果表明,这种建模方法使计算和存储大为简便,可为设计人员提供设计参考。     

柯布──道格拉斯模型在喷灌经营管理中的应用

本文以柯布──道格拉斯生产函数模型，研究分析喷灌经营管理运行中诸生产要素与总效益之间的关系，这为评估喷灌的运行管理现状及水平是否适应低耗高质高效的要求，提供了研究方法。     

车辆防抱制动系统滑移率自适应控制策略

分析了车辆制动过程的数学模型,为了克服模型非线性、变参数对系统的影响,采用了参考模型自适应控制.为了适应各种路况条件,使车辆以比较理想的滑移率制动,缩短制动时间,提出了滑移率自适应调节方法.通过仿真证明,该方法简单可行,控制效果好.     

Wood Production Potentials of Fenno-Scandinavian Forests Under Nature-Orientated Management

In Finland, Norway and Sweden forest management is presently changing towards a more nature-orientated management. In this study the European Forest Information SCENario (EFISCEN) model was applied to determine how this change might affect the potential for wood production in the three countries. Three different management regimes, traditional, traditional with nature conservation (''conservation''), and longer rotations with nature conservation (''conservation +''), were combined with two alternative felling levels. The results show that conserving 6.1-8.8% of the older forests in the southern regions had no limiting effects on production levels, as foreseen by the European Timber Trend Studies V by the UN-ECE for Finland and Sweden. Under the conservation + scenario, maximum sustainable felling levels decreased to 84, 79 and 72% of the present levels in Finland, Norway and Sweden, respectively. Increasing the rotation length put more pressure on the older age classes and thus did not increase the average age of the forest. If the consumption of wood increases as quickly as indicated by other studies, it will be hard to fulfil that demand and at the same time conserve considerable areas of forest in the southern regions of the countries.