基于粒子群算法的混联机构神经网络自适应反演控制

针对含有不匹配干扰的混联机构轨迹跟踪控制问题，提出了一种极限学习机与自适应反演控制相结合的控制策略。在对干扰进行分析的基础上，分别采用两个极限学习机网络对系统中的匹配和不匹配干扰进行逼近和补偿。基于Lyapunov函数稳定性设计了混联机构的控制律与自适应律，实现混联机构的轨迹跟踪控制。由于控制器可调参数较多，采用粒子群算法进行控制器参数的寻优整定。仿真结果表明，所提出的控制方法具有良好的轨迹跟踪精度和鲁棒性。     

采伐干扰下次生林灌木层主要树种的生态位动态

为探讨采伐干扰下的灌木层主要树种的生态适应性,用Levins、Schoener和Pianka等生态位测度公式定量分析闽北天然次生林采伐后灌木层主要树种生态位动态。结果表明:样地中黄绒润楠和刺毛杜鹃的生态位最大,属优势树种;甜槠、石栎和映山红为衰退树种,在林分生产作业中需加以保护。采伐干扰促进了灌木层主要树种的生态位释放,各树种的生态位宽度在伐后10~15 a达到峰值,随后下降,伐后20 a接近于未采伐林的生态位;各树种弱度和中度择伐的生态位宽度值约占其生态位宽度总值的50%。生态位重叠值在伐后10 a最大,随后下降,到伐后20 a趋于稳定状态。因此灌木层主要树种生态位的最优期为伐后10~15 a,最佳采伐干扰强度为弱度和中度择伐。     

基于终端滑模和扰动观测的Buck型变换器复合控制技术

目前功率变换器的控制方法以纯状态反馈为主，当干扰较大时一般需要通过较大的增益来抑制干扰的影响，而高增益控制器通常会影响闭环系统的动态或稳态性能。为进一步提高Buck型变换器控制系统的动态和稳态性能，提出了一种基于非奇异终端滑模控制方法和扰动观测技术的复合控制方案。首先，考虑外界干扰、系统不确定以及参数变化的影响，建立Buck型功率变换器的平均状态模型；其次，基于非奇异终端滑模控制方法，设计滑模控制器，实现Buck型变换器的基本电压调节功能；最后，利用扰动观测技术，构造非线性扰动观测器实现对扰动的观测，并将扰动观测值作为前馈与状态反馈结合形成复合控制，进一步改善了系统的性能。仿真和实验验证了所提方法的有效性。     

Disturbance ecology benchmarks for Schima–Castanopsis forests in the central Himalayas

Roughly 2.8 billion people burn wood for basic energy needs, and traditional wood-fuel represents ~55% of global wood harvest. With increasing anthropogenic disturbance of natural forests, the “stability/fragility” paradigm of forest ecology is gradually being replaced by a “disturbance/recovery” paradigm. In order to understand effects of human-induced disturbances on natural forest ecosystems, and to plan for recovery of disturbed forests, appropriate metrics become necessary. Such metrics will aid in assessment and management of forests for carbon sequestration, biodiversity conservation, ecosystem health, and sustainability of natural resources. Such metrics are especially needed in “wood-fuel hotspots” of the world where over 275 million people live and harvest wood-fuel unsustainably. In this article, I provide metrics of human-induced disturbance in Nepal’s Schima–Castanopsis dominated forests and show relationships of disturbance intensity with forest structure and composition, site productivity potential, natural regeneration, and tree species diversity. Benchmark data were collected from survey of two protected reference forests and compared against three other forests representing a disturbance gradient. The Schima–Castanopsis association is a common dominant forest type in the warm temperate zone of the central Himalayas, and the findings from this study should have wider application.     

Fish functional diversity responses following flood pulses in the upper Paraná River floodplain

Flood pulses are the main force driving the dynamics of aquatic communities in floodplains. The responses of communities to environmental changes following flood pulses usually demand a time lag to appear and reach the climax. We assembled a data set of 16 years of fish samplings to assess the relationship between water level and four functional diversity measures, in the upper Paraná River floodplain. Specifically, we approached four aspects of each relationship between water level and functional diversity: nature (positive or negative), sensitivity (response intensity), responsiveness (response delay) and extent (response duration). The nature of the relationship between water level and functional diversity was positive in all cases. Functional richness (FRic) responded right after flood pulses, although with shorter extent. Abundance‐dependent functional measures (evenness—FEve; divergence—FDiv; and Rao's quadratic entropy—Rao's Q) presented delayed responses, reaching peaks more than 1.5 years after flood pulses. Significant effects of floods on fish functional diversity were observed for more than 3 years, although the highest functional diversity was observed with 1.8 years, on average. More importantly, flood pulses had no longer significant effects on functional diversity after 4 years. Regarding conservation strategies in regulated systems, flood events should occur every 2 or 3 years, with adequate timing (October‐November), intensity (up to 450 cm) and duration (at least 50 uninterrupted days). Intervals longer than 3 years or inadequate timing, intensity and duration could dramatically decrease functional diversity and compromise ecosystem services.     

Responses of Australian sea lions,Neophoca cinerea,to anthropogenic activities in the Perth metropolitan area,Western Australia

1. Tourist‐based activities, partly due to their rapid increase, have raised concerns regarding the impacts of anthropogenic activity on marine fauna. Documented effects on pinnipeds in proximity to humans include changes in behaviour, site use and potentially higher aggression levels towards people. Effects vary considerably between populations and sites, thus requiring separate assessment of human impacts on activity and energy budgets.
2. Responses of the endangered Australian sea lion, Neophoca cinerea, to human visitation were recorded from November 2013 through April 2014. Exposure levels and response types to anthropogenic activities were assessed at two easily accessible locations with different management schemes, Seal (landing prohibited) and Carnac (landing permitted) islands, Western Australia. Exposure levels were measured as both stimulus type (i.e. ‘People’, ‘Paddlers’, ‘Small’, ‘Medium’, and ‘Large vessels’, ‘Tour vessels’, and ‘Jet skies’), and people (‘Direct’, ‘Attract’, ‘Interact’, ‘View’, ‘Incidental’, ‘Water’, ‘Low‐level’), and vessel activities (‘Interact’, ‘Approach/Follow’, ‘Anchor noise’, ‘Engine noise’, ‘Close to beach’, ‘Moderate/Fast travel’, ‘Slow travel’, ‘Transit’, ‘Drift/At anchor’, ‘Aircraft noise’).
3. Exposure levels varied significantly between the islands in numbers, stimuli type, duration and minimum approach distances. The instantaneous behaviours of ‘Lift head’, ‘Interact’ and ‘Sit’ were the most frequent responses. ‘Aggressive’ and ‘Retreat’ responses, the highest disturbance levels measured, occurred on Carnac approximately once per day, but rarely on Seal Island. ‘Aggressive’ behaviour towards ‘People’ was observed only on Carnac Island and elicited only by ‘People’. ‘People’, ‘Tour vessels’, and scenic ‘Aircrafts’ on both islands as well as ‘Jet skis’ on Carnac Island had the highest probability of triggering responses. Owing to their relatively high visitation at Seal Island, ‘Paddle powered vessels’, followed by ‘Tour vessels’ elicited the highest number of responses, compared with ‘People’, ‘Small’, and ‘Medium vessels’ at Carnac Island. The majority of responses occurred when any stimulus type was at short‐range (≤10 m), and ‘People’ ‘Viewing’ N. cinerea elicited most. Vessels triggered more responses at larger ranges than ‘People’.
4. To limit close‐range access to N. cinerea, one possibility is to close the beach at Carnac Island to human visitation and increase the minimum approach distance by vessels and ‘People’ by installing marker buoys at least 15 m from the shore.

© 2016 The Authors. Aquatic Conservation: Marine and Freshwater Ecosystems published by John Wiley & Sons, Ltd.     

A network meta‐analysis of threats to South American fish biodiversity

While environmental alterations have made Homo sapiens the hyperkeystone species of the globe, biotic homogenization initiated a new era, the “Homogenocene.” Still, some terrestrial and aquatic ecosystems in South America are considered pristine and wild, which can lead to a general faith that economic progress is consistent with conservation strategies, even without scientific support. We compiled anthropogenic threats to fish biodiversity in a hierarchical meta‐analysis, along with an evidence synthesis of threats related to biological invasions, based on peer‐reviewed research with the aim to represent the actual conservation status of the South American ichthyofauna. We highlighted human‐related threats and synergistic effects of biological invasions, climate change, environmental alterations (e.g., pollution, aquaculture and damming) and fisheries. Considering measures that reinforce novel alien fish (e.g., artificial hybrids or genetically modified) introductions, it became clear why an eventual increase in local or regional species richness is not always beneficial to aquaculture, biodiversity, human well‐being or nature. In fact, citizens in all societal roles, including scientists, should revise their concepts about threats to fish biodiversity. Environmental policies require more than taxonomic diagnostics to achieve conservation goals under an incompatible scenario of a multiplying number of fish species and biotic homogenization. We advocate for countries in South America using science‐based strategies useful to maintain their social and economic growth along with their “remaining nature.” We live a crucial moment when the government overlooks threats to biodiversity and uses agribusiness as the most acceptable manner of fuelling the economy.     

火干扰对喀纳斯泰加林土壤有机碳的影响

为了研究不同火干扰烈度对喀纳斯泰加林土壤有机碳含量及碳储量的影响,以喀纳斯泰加林同一火烧时间不同火干扰烈度的火烧迹地为研究对象,对不同土层(0~10 cm,10~25 cm)有机碳含量进行测定并估算其碳储量,采用单因素方差分析及LSD多重比较进行差异显著性分析。结果表明,相同火干扰烈度下0~10 cm土层土壤有机碳及碳储量均大于10~25 cm土层。0~10 cm土层中碳储量表现为：未火烧>轻度火干扰>中度火干扰>重度火干扰,在10~25 cm土层中碳储量表现为：轻度火干扰>中度火干扰>未火烧>重度火干扰。0~10 cm土层的土壤有机碳含量和碳储量以及10~25 cm土层的有机碳含量损失程度会随着火干扰烈度的增加而增加。10~25 cm土层碳储量会在轻度和中度火干扰后出现增加。     

耕地植被人为干扰格局动态变化特征及其尺度效应

为探究耕地植被人为干扰格局动态变化特征及其尺度效应,该研究以黑龙江省大庆市为研究区,通过遥感与地理信息系统相结合的方法计算和识别耕地植被人为干扰,分析5个时段(1990-1995年、1995-2000年、2000-2005年、2005-2010年和2010-2016年)耕地植被人为干扰格局动态变化特征;并采用滑动格窗法和ISODATA聚类算法,对耕地植被人为干扰格局的尺度效应进行了研究。结果表明:1)当耕地植被受到的人为干扰增多时,研究区耕地植被人为干扰的分布趋向于集中;反之则趋向于分散。2)研究期内占耕地植被人为干扰面积50%以上的地区的尺度效应并不明显,但仍有占面积11.81%～18.33%的地区人为干扰的尺度效应十分显著,这些地区是人为干扰分布最集中和最不均匀的地区。3)在耕地植被人为干扰尺度效应显著地区中,土地利用类型未发生变化的地区所占的面积比例呈上升趋势,说明耕地保护政策产生了较好的效果;但耕地植被人为干扰的主要来源是农业生产行为,应将未来政策关注的重点转为对耕地土壤理化性质稳定性的维护。4)从多尺度空间聚类类型可以看出,在大部分尺度效应显著地区,尺度越小耕地植被人为干扰越显著,说明发展适度规模农业是降低人为干扰、提高粮食产量的有效途径。研究对于实现耕地保护和粮食稳定生产具有重要意义。     

Disturbance‐induced responses of VHF and satellite tagged harbour seals

1. The response of individual harbour seals (Phoca vitulina) to controlled and sporadic disturbances when hauled out in the Anholt seal reserve, Denmark, was studied. Sporadic disturbances from pedestrians, boats, low‐flying aeroplanes and grey seals (Halichoerus grypus) were observed in or near the reserve. VHF and satellite transmitters were attached to eight harbour seals to determine post‐disturbance and undisturbed behaviour during the pre‐breeding period (25 April to 21 May 2008).
2. Both disturbed and undisturbed seals mostly returned to the haul‐out site from dusk and throughout the night. Only pedestrian disturbances caused significantly longer‐lasting trips at sea than undisturbed trips, while the other types of disturbance did not affect trip duration. The general maximum extent and area‐use of disturbed and undisturbed trips were comparable and almost all at‐sea locations were within 40 km from the haul‐out site. The maximum extent of post‐disturbance trips, however, varied among individuals and disturbance types, and was strongly correlated with the duration of trips.
3. Disturbed and undisturbed seals used the same areas, suggesting that these areas represent normal foraging areas. This may indicate that harbour seals reduce the cost of being disturbed by foraging after disturbances instead of waiting close to the haul‐out site to resume hauling out.
4. During the pre‐breeding period the seals showed a very high site fidelity by consistently returning to the same haul‐out site, even when subjected to repeated disturbances.

Copyright © 2013 John Wiley & Sons, Ltd.