基于BP神经网络的人机联合率定SWMM研究

为解决雨洪管理模型(SWMM)在率定过程中参数复杂、过程繁琐等问题。以西宁市某地块为例建立SWMM模型,利用Morris筛选法进行参数灵敏度分析,并根据灵敏度分析的结果进行人工率定;另外利用BP神经网络对模型进行率定,并结合参数灵敏度对其进行优化。对3种率定方案进行分析,结果表明:水文水力模块参数的相对灵敏度基本一致,其中灵敏度较大的参数为子汇水区面积(Area)、不透水率(Imperv)和不透水区洼地蓄积量(Destore-Imperv),并且不同降雨条件下模型参数的灵敏度存在差异。经过优化后的BP神经网络参数率定方法的模型模拟效果最好,纳什系数最大。结合灵敏度优化BP神经网络的人机联合率定方法一方面能提高BP神经网络率定的准确性,另一方面又能提高传统人工率定的效率。     

Prediction of baking results from farinograph measurements by using stepwise linear regression and artificial neuronal networks

In an effort to extract additional data from farinograph experiments a model was developed to simulate the measurements and correlate the parameters of the model with results from baking tests. This additional information can be used in bakeries to predict the baking properties of the flours and adjust the recipes to maintain a constant product quality. For this eight different flours were characterized with a farinograph and 13 different results from baking experiments. An approach with five nonlinear differential equations was able to model the farinograph measurements very well (average R² = 0.995 ± 0.005). While a stepwise multilinear regression only showed weak correlations in cross validation between a single parameter of the model and the baking volume (R² = 0.745) and the volume yield (R² = 0.796) respectively, the artificial neuronal network was more successful. For the baking weight (R² = 0.926), the dough yield gross (R² = 0.909) and net (R² = 0.913) strong correlations were found. A good correlation for the baking volume (R² = 0.853) was also determined, while the volume yield showed comparable results to the linear regression (R² = 0.792).     

海贝柱自动晾晒设备设计

针对当前传统人工晾晒海贝柱效率低下、容易造成污染等问题。设计了海贝柱自动晾晒机械,采用阶梯式晾晒生产线对湿海贝柱分级晾晒,多条晾晒生产线分时、高效的工作,机器上方设有遮板以保证晾晒作业不被污染而且不受天气条件的影响。整机采用stm32单片机、传感器等技术实现自动化作业,机器控制系统监测到生产线发生故障或海贝柱出现堆积时会立即停止工作以便于检查故障、减少损失。试验结果表明:机械晾晒不会造成海贝柱内的营养成分的流失,机械晾晒海贝柱的干基含水率达到30%仅需12h,机械晾晒海贝柱能节省大量的干燥时间,显著的提高了晾晒效率,避免了人工晾晒过程导致的二次污染和损失。该研究为海贝柱自动晾晒机械的研制提供了技术参考。     

人工神经网络在电力系统短期负荷预测中的应用

根据电力系统短期负荷变化的特性，提出BP模型在实际负荷预测应用中的方法和步骤．对BP网络结构、样本空间、收敛性等作了有针对性的研究．结果表明：多层神经网络应用于电力系统短期负荷预测是可行和有效的．其预报结果比传统的负荷预测方法更准确、经济、效果更好．     

Evaluation of a combined penetrometer for simultaneous measurement of penetration resistance and soil water content

A combined penetrometer is an appropriate tool to measure the soil cone resistance and the water‐content profile. As a relatively new technique, a combined capacitance‐penetrometer for the simultaneous measurement of cone index and soil water content was developed at the Department of Agricultural Engineering of Bonn University in 2002. The objective of this study was the evaluation of the effectiveness and applicability of the innovated penetrometer with a focus on three aspects: (1) A capacitance sensor with two electrode configurations was calibrated for silt loam, sandy loam, and sand. The calibration results show that both electrode configurations have sufficient water‐content sensitivity, but soil‐specific calibrations seem necessary. (2) Under laboratory conditions, the dynamic resolution and response of the capacitance‐penetrometer were validated, and its radius of influence was determined. (3) The field measurement results demonstrate that this measurement technique can be used to improve the interpretation quality of soil cone index data.     

木材干燥应力连续监测方法的研究

本文在卡普法的基础上，提出了木材干燥应力连续监测方法，并设计了所需的配套装置。该方法采用涡流传感器对干燥过程中卡普片的矢高进行连续测试，同时通过计算机自动采集、存储数据。结果表明：矢高与干燥时间存在精密的相关性，它们之间的相关系数大部分达到0．99以上，从而实现了木材干燥应力的自动与连续监测。     

Compatible measurements of volumetric soil water content using a neutron probe and Diviner 2000 after field calibration

Field calibrations for a neutron probe and a capacitance sensor (Diviner 2000) for measuring the soil water content of a shrinking–swelling clay soil were substantially different from commonly used default values. Using our field calibrations, the two instruments estimated similar changes in the cumulative water content of a soil profile (0–1 m depth) over one growing season.     

电阻电桥传感放大器性能的改善

在构成基本型电阻电桥传感放大器的基础上给出了三种改善其性能的方法。１）补偿法。２）补偿加放大器改进综合法。３）负反馈法。通过实验证明了理论分析和实际效果是一致的，从而对电阻传感器在非电量检测中的广泛应用具有一定的实际意义。     

电流模式神经网络

把电流模式电路引进了神经网络理论，提出了一种用电流模式电路构成的神经网络阈值处理单元，并由此构成了电流模式Ｈｏｐｆｉｅｌｄ神经网络，此网络中信号都是以电流的形式出现的。     

双板差分冲量式谷物流量传感器设

设计了一种双板冲量式谷物流量传感器及其差分消振电路,以消除车身振动对测产精度的影响.振动试验表明,在没有谷物输入时联合收获机车身振动导致前后2个检测板产生的2个振动信号是一对共模信号,可通过电路差分的方法消除;差分后传感器的输出电压的信噪比由2.5dB提高到62dB.在零喂入量情况下,振动产生的零点误差小于0.01g/s;在联合收获机正常作业的喂入量范围内(0.5～2kg/s),实验室测量误差小于3%.田间实际测量误差不大于5%.