基于粒子群算法优化BP神经网络的土壤含水量短期预测模型

为提高土壤含水量预测精度,基于物联网监测数据,提出了粒子群算法(PSO)优化BP神经网络的土壤含水量预测方法。首先应用主成分分析法筛选出影响土壤含水量的关键影响因子,然后构建8-5-1的BP神经网络拓扑结构,应用粒子群算法优化BP神经网络的初始权值和阈值。结果表明:与传统BP神经网络相比,新模型优化了网络结构,避免了陷入局部最优解,具有良好的预测效果;模型的评价指标平均绝对误差、平均绝对百分误差、误差均方根分别为0.259 2、0.010 5和0.135 6,与单一BP神经网络相比,预测精度更高,可满足实际的土壤含水量预测的需要。     

基于热致液晶弹性体的仿生心脏泵的设计

为设计一款仿生心脏功能较佳的心室泵,采用柔性、热致伸缩的热致形变弹性体代替心肌,根据螺旋心室心肌带理论,选择心肌带的左心室段,并选取了Hilbert平面填充曲线作为热致导体在热致形变弹性体中的走行方式,设计出与真实心脏类似的心室泵结构.通过试验测量了所设计的心室泵的射血分数与心率,结果显示该心室泵的搏出量高于自然心脏,而其心率远低于自然心脏.相对于机械泵,所设计的心室泵结构与真实心脏类似,具有收缩能力强,射血分数高的优点,且不易破坏血细胞,生物相容性较好.不足之处在于,热致动液晶弹性体变形响应慢,暂时还未能达到正常心脏的心率.论证了构建双稳态结构,提高仿生心室泵效率的可能性,以及对于心肌纤维走向较为复杂的其他心肌带段,双轴拉伸机械雕刻对其仿生的可能性.此研究可为容积心脏泵的研发提供新的思路,也可以对右心衰竭等心脏疾病起到辅助治疗的作用.     

基于PLSR-BP复合模型的红壤有机质含量反演研究

对红壤地区土壤有机质进行快速预测,以满足智慧农业与精准施肥的需要。以江西省奉新县北部为研究区域,采用1 km×1 km标准格网划分研究区进行采样,共得到红壤样本248个。对土壤光谱进行了包含分数阶导数在内的3种数学变换方法,将经过P=0.01显著性检验的波段用于模型的构建,选用偏最小二乘回归(PLSR)和BP神经网络建立土壤有机质含量预测模型。结果表明:当对红壤光谱数据进行1.5阶导数变换后再使用PLSR-BP复合模型对土壤有机质含量进行预测时的结果为最优,训练集R~2=0.89,RMSE=4.68g·kg~(-1),验证集R~2=0.87,RMSE=5.55g·kg~(-1),RPD=2.75。1.5阶导数对红壤光谱数据的变换能够更好地突出与有机质相关的特征信息,有助于其含量预测。PLSR-BP复合模型预测精度优于单一模型,能够较好地预测红壤有机质含量,为精准农业快速监测红壤有机质含量提供了新的途径。     

圆柱镂空型人工鱼礁波流水动力特性数值模拟

研究人工鱼礁在波流作用下的水动力特性,对于人工鱼礁的设计具有重要的意义。基于有限体积法,采用边界造波,利用自由表面捕捉法(VOF)捕捉自由水面,建立了可以分别模拟纯波、均匀流以及波流共同作用下人工鱼礁水动力特性的多功能三维数值波流水槽。基于该数值模型对不同波流工况作用下圆柱型镂空人工鱼礁水动力特性进行数值模拟,并与物理模型试验结果进行比较。结果显示,人工鱼礁数值模拟受力与模型试验结果吻合良好,人工鱼礁所受的波流力最大值随着波高、周期和水流流速的增大而增大;人工鱼礁处于波流场波峰正下方时,背涡流的面积随着水流流速的增大而增大,随着波高、周期增大而减小。对单独均匀流作用、单独波浪作用和波流联合作用下人工鱼礁的水动力特性对比研究表明,人工鱼礁所受的最大波流力比最大波浪力、水流力都大,波流联合作用下的流场效应最显著,在礁体的后部形成了较大规模的漩涡结构。     

Potential application of baited remote underwater video to survey abundance of west coast rock lobster Jasus lalandii

The first assessment of the baited remote underwater video system (BRUVs) for monitoring rock lobster Jasus lalandii (Milne‐Edwards) in South Africa was compared to annual ring net and trap survey data. Count data from 58 BRUVs, 36 ring net and 95 trap samples were fitted with generalised additive models using depth, habitat, time and spatial gradient as independent variables. A power analysis was used to determine the number of samples needed per annum to detect a 5 and 10% increase and decrease in population size. The BRUVs outperformed the nets and traps in terms of sampling effort and were better at detecting changes in abundance of lobster than abundance of three fish species. Considering factors such as cost, analysis time, additional data and environmental impact, traps are the most feasible method in deeper waters but BRUVs could replace nets and traps in depths less than 60 m, low‐density areas and marine reserves.     

Artificial neural network modeling of microbial community structures in the Atlantic Forest of Brazil

Microbial communities vary across the landscape in forest soils, but prediction of their biomass and composition is a difficult challenge due to the large numbers of variables that influence their community structures. Here we examine the use of artificial neural network (ANN) models for extraction of patterns among soil chemical variables and microbial community structures in forest soils from three regions of the Atlantic Forest of Brazil. At each location, variations in soil chemical properties and FAME profiles of microbial community structures were mapped at 20 × 20 m intervals within 10 ha parcels. Geostatistical analyses showed that spatial variability in soil physical and chemical variables could be mapped at scale distances of 20 m, but that FAME profiles representing the microbial communities were highly variable and had no spatial dependence at the same scale in most cases. RDA analysis showed that FAME signatures representing different microbial groups were positively associated with soil pH, OM, P and base cations concentrations, whereas microbial biomass was negatively associated with the same environmental factors. In contrast, ANN models revealed clear relationships between microbial community structures at each parcel location, and generated verifiable predictions of variations in FAME profiles in relation to soil pH, texture, and the relative abundances of base cations. The results suggest that ANN modeling provides a useful approach for describing the relationships between microbial community structures and soil properties in tropical forest soils that were not able to be captured using geostatistical and RDA analyses.     

Induction of meiotic gynogenesis in bagrid catfish (Pseudobagrus ussuriensis) with homologous sperm and its confirmation for female homogamety

The bagrid catfish, Pseudobagrus ussuriensis, exhibits significant sexual dimorphism in growth rate and body size with males growing faster than females. Therefore, an all‐male culture can dramatically increase the output and profitability of fishery products. According to the monosex breeding route, super‐male individuals’ acquirement by XY male sex reversal and artificial gynogenesis is the key step. An effective protocol to induce meiotic gynogenesis using homologous sperms has been developed in this study. The most effective UV irradiation for sperm genetic inactivation was found to be at a distance of 20 cm with 66 μW/cm² light intensity for 25 min. Optimal treatment for cold shock was 5 min post‐fertilization at 0‐4°C for 30 min, which gave the best survival rate of 13.65 ± 2.87%. The sex ratio in the meiotic gynogens showed a significant female‐biased deviation (p < .05); thirty meiogynogens and their parents were further analysed using a male‐specific AFLP marker, of which only the male parent produced a male‐specific DNA band of 412 bp. These results indicated the female homogametic XX/XY sex determination system in P. ussuriensis. The optimization of a protocol for the successful induction of meiogynogenesis in the bagrid catfish lays the basis for all‐male production and is useful in ascertaining the genetic sex determination system in this promising aquaculture species.     

Differences in the effectiveness of reproduction among ten breeding strains of the common carp Cyprinus carpio L. after stimulation of ovulation with carp pituitary homogenate or [(D‐Ala6, Pro9NEt) mGnRH‐a + metoclopramide] (Ovopel)

This study focused on the reproduction effectiveness of 10 breeding strains of common carp (Polish strains 2, 3, 6; Hungarian strains 0, W, 7; Lithuanian strain B; French strain F; Israeli strain D and Yugoslavian strain J) after stimulation of ovulation with carp pituitary homogenate or Ovopel. The percentage of females with recorded ovulation became higher after Ovopel treatment in as many as eight breeding strains (2, 3, 6, 0, W, B, F, D). The interaction between the spawning agent and the breeding strain was statistically significant (P ≤ 0.05) for the mass of spawn and non‐significant for traits determining the quality of eggs. Among the 10 breeding strains, in 4 (3, 7, B, J) the least‐squares means (LSM) for the weight of egg (g) were higher for fish stimulated with Ovopel. Among the strains tested, in seven (3, W, F, J, 2, 7, B), the LSM for the percentage of living embryos after 36 h incubation were higher for fish treated with Ovopel. The evaluation of reproduction effects based on the values of such parameters as the percentage of ovulating females, weight of eggs (g) and the percentage of live embryos (36 h) showed that after Ovopel treatment the poorer effects were obtained only for strains 6, 0 and D.     

基于改进VMD的离心泵空化声发射信号特征提取

针对变分模态分解算法中分解层数和惩罚因子不易确定的问题,提出一种改进变分模态分解(improved variational mode decomposition,IVMD)算法,并将其应用于离心泵空化声发射信号特征提取.应用IVMD算法时,首先根据包络熵差异系数确定变分模态分解的分解层数;然后采用人工蜂群算法优化得出惩罚因子,并将其作为变分模态分解的最佳输入参数.利用IVMD算法对仿真信号进行分析,并与集合经验模态分解结果进行比较.以60%额定流量下采集到的离心泵进口处的声发射信号为例进行IVMD计算,分析携带原信号大量信息的信号分量的频域特征及其绝对能量随离心泵空化状态变化的关系.结果表明:IVMD算法能够择优确定分解层数和惩罚因子,实现非平稳信号的自适应分解.反映离心泵空化状态的声发射信号特征频率集中在50,100 kHz及其附近.随着离心泵空化从无到有、从弱到强的变化,这2个特征频率范围信号分量绝对能量值呈“基本保持不变-减小-增大”的变化规律.