Bird detection and species classification with time‐lapse images around a wind farm: Dataset construction and evaluation

Collisions of birds, especially endangered species, with wind turbines is a major environmental concern. Automatic bird monitoring can be of aid in resolving the issue, particularly in environmental risk assessments and real‐time collision avoidance. For automatic recognition of birds in images, a clean, detailed, and realistic dataset to learn features and classifiers is crucial for any machine‐learning‐based method. Here, we constructed a bird image dataset that is derived from the actual environment of a wind farm and that is useful for examining realistic challenges in bird recognition in practice. It consists of high‐resolution images covering a wide monitoring area around a turbine. The birds captured in these images are at relatively low resolution and are hierarchically labeled by experts for fine‐grained species classification. We conducted evaluations of state‐of‐the‐art image recognition methods by using this dataset. The evaluations revealed that a deep‐learning‐based method and a simpler traditional learning method were almost equally successful at detection, while the former captures more generalized features. The most promising results were provided by the deep‐learning‐based method in classification. The best methods in our experiments recorded a 0.98 true positive rate for bird detection at a false positive rate of 0.05 and a 0.85 true positive rate for species classification at a false positive rate of 0.1.     

海上风电场对鸟类的影响及其危害预防

[目的]风能作为公认的最成熟的可再生能源技术之一,近年来发展迅速.中国在近年开发了大量的海上风电项目.但风电场对环境,尤其是对鸟类的影响引发了人们大量的担忧和研究.在碳达峰实现之前,我国的海上风能产业将持续增长,需要全面了解风电场对鸟类的影响.[方法]通过总结现有研究,对海上风电场引起的鸟类问题进行了综述,并讨论了可将...     

基于运行数据的风电机组转矩控制性能评估

风电机组的性能评估方法具有多样性及复杂性的特点,基于风电场SCADA系统中采集的大量风电机组运行数据,对风电机组转矩控制的性能评估方法进行了研究。在深入分析风电机组中发电机转速与发电机转矩关系的基础上,提出了风电机组在最佳风能利用系数Cp(max)跟踪区内的转矩优化控制的性能评估方法。通过筛选有效数据,拟合计算出风电机组的实际运行转矩增益系数;再通过与理论最优转矩增益系数进行对比,找出风能捕获能力较弱的风电机组,进而采取措施提高其发电量。通过软件仿真及案例分析表明,该方法在不增加设备及成本的情况下,可有效识别因转矩控制的性能差而影响发电量的风电机组,以便及时进行控制策略调校,维护风电场的利益。     

集团型风电企业综合联网监控管理系统的构建

以五大发电集团为代表的集团型风电企业已经成为我国风电投资开发的主体力量,如何管控好数量急剧增加的风电场成为企业安全生产中重要的课题。从风电场安全视频监控的角度,针对其普遍实行的三级责任主体的管控模式,提出了一种综合联网监控管理系统的构建方案。从需求分析、系统设计思路、系统的详细设计3个方面进行了阐述,其中专业网络视频管理系统是整个系统的核心,要求其具有容量大,承载能力强,可靠性好,扩展性、兼容性强的特点,且能容纳业已存在的若干风场的视频监控系统。文中提出的方案对集团型风电企业的综合联网监控管理系统的构建具有借鉴作用。     

Development of a modified stochastic subspace identification method for rapid structural assessment of in‐service utility‐scale wind turbine towers

The strong drive to harness wind energy has recently led to rapid growth of wind farm construction. Wind turbine towers with increased sizes and flexibility experience large vibrations. Structural health monitoring of wind turbines is proposed in the wind energy industry to ensure their proper performance and save maintenance costs. This study proposes a system identification method for vibration‐based structural assessment of wind turbine towers. This method developed based on the stochastic subspace identification method can identify modal parameters of structures in operating conditions with harmonic components in excitations. It benefits wind turbine tower structural health assessment because classical operational modal analysis methods can fail as periodic rotation excitation from a turbine introduces harmonic disturbance to tower structure response data. The effectiveness, accuracy and robustness of the proposed method were numerically investigated and verified through a lumped‐mass system model. The method was then applied to an in‐service utility‐scale wind turbine tower. The field testing campaign and modal parameter identification as well as structural assessment results were presented. Copyright © 2017 John Wiley & Sons, Ltd.     

Wind farm layout optimization using area dimensions and definite point selection techniques

Wind turbines are the biggest rotating machines on earth, operating in the lowest part of the earth boundary layer. Designing the layout scheme of wind farms is a challenging job to researchers, as there are many design objectives and constraints due to the multiple wake phenomena. This paper proposes an area rotation method to find the optimum dimensions of the wind farm shape, where maximum area could face the free stream velocity. Afterwards, a novel method called Definite Point Selection (DPS) is developed to place the turbines in order to operate at maximum, while providing obligatory space between adjacent turbines for operation safety. This method can be used to identify the zero wake effect points at wind farm. The result from this study shows that the proposed method is more effective to increase the overall power of a wind farm than the previous methods. Also, the power output of the wind farm by using combined area rotation and DPS methods was increased even when using the same number of wind turbines.     

Convolutional neural network framework for wind turbine electromechanical fault detection

Effective and timely health monitoring of wind turbine gearboxes and generators is essential to reduce the costs of operations and maintenance activities, especially offshore. This paper presents a scalable and lightweight convolutional neural network (CNN) framework using high-dimensional raw condition monitoring data for the automatic detection of multiple wind turbine electromechanical faults. The proposed approach leverages the potential of combining information from a variety of signals to learn features and to discriminate the types of fault and their severity. As a result of the CNN layers used to extract features from the signals, this architecture works in the time domain and can digest high-resolution multi-sensor data streams in real-time. To overcome the inherent black-box nature of AI models, this research proposes two interpretability techniques, multidimensional scaling and layer-wise relevance propagation, to analyse the proposed model's inner-working and identify the signal features relevant for fault classification. Experimental results show high performance and classification accuracies above 99.9% for all fault cases tested, demonstrating the efficacy of the proposed fault-detection system.     

Ground flora,small mammal and bird species diversity in miscanthus (Miscanthus×giganteus) and reed canary-grass (Phalaris arundinacea) fields

Wildlife monitoring of two miscanthus and two reed canary-grass fields in Herefordshire, England was carried out in 2002, 2003 and 2004 to investigate the ecological impact of perennial biomass grass crops on ground flora, small mammals and birds. Quadrats were used to record percentage ground vegetation cover within and around the periphery of each crop. Small mammals were sampled by live trapping using Longworth traps. The common bird census technique was used to monitor populations of birds. Miscanthus fields were richer in weed vegetation than reed canary-grass or arable fields. Bird use of the biomass crop fields varied depending on species. There were considerably more open-ground bird species such as skylarks (Alauda arvensis), lapwings (Vanellus vanellus) and meadow pipits (Anthus pratensis) within miscanthus than within reed canary-grass fields. There was no particular crop-type preference by the small mammal species, but rather a preference for good ground cover and little land disturbance, which was provided by both biomass crops. Ground flora, small mammals and most of the bird species (except open-ground birds) were found more abundantly within field margins and boundaries than in crop fields indicating the importance of retaining field structure when planting biomass crops. The miscanthus work relates entirely to young crops, which may be representative of part of the national crop if large areas are cultivated for rhizomes. The findings from the current project indicate that perennial biomass grass crops can provide substantially improved habitat for many forms of native wildlife, due to the low intensity of the agricultural management system and the untreated headlands.     

Modeling of the dynamics of wind to power conversion including high wind speed behavior

This paper proposes and validates an efficient, generic and computationally simple dynamic model for the conversion of the wind speed at hub height into the electrical power by a wind turbine. This proposed wind turbine model was developed as a first step to simulate wind power time series for power system studies. This paper focuses on describing and validating the single wind turbine model, and is therefore neither describing wind speed modeling nor aggregation of contributions from a whole wind farm or a power system area. The state‐of‐the‐art is to use static power curves for the purpose of power system studies, but the idea of the proposed wind turbine model is to include the main dynamic effects in order to have a better representation of the fluctuations in the output power and of the fast power ramping especially because of high wind speed shutdowns of the wind turbine. The high wind speed shutdowns and restarts are represented as on–off switching rules that govern the output of the wind turbine at extreme wind speed conditions. The model uses the concept of equivalent wind speed, estimated from the single point (hub height) wind speed using a second‐order dynamic filter that is derived from an admittance function. The equivalent wind speed is a representation of the averaging of the wind speeds over the wind turbine rotor plane and is used as input to the static power curve to get the output power. The proposed wind turbine model is validated for the whole operating range using measurements available from the DONG Energy offshore wind farm Horns Rev 2. Copyright © 2015 John Wiley & Sons, Ltd.     

Grassland bird response to harvesting switchgrass as a biomass energy crop

The combustion of perennial grass biomass to generate electricity may be a promising renewable energy option. Switchgrass (Panicum virgatum) grown as a biofuel has the potential to provide a cash crop for farmers and quality nesting cover for grassland birds. In southwestern Wisconsin (near lat. 42°52′, long. 90°08′), we investigated the impact of an August harvest of switchgrass for bioenergy on community composition and abundance of Wisconsin grassland bird species of management concern. Harvesting the switchgrass in August resulted in changes in vegetation structure and bird species composition the following nesting season. In harvested transects, residual vegetation was shorter and the litter layer was reduced in the year following harvest. Grassland bird species that preferred vegetation of short to moderate height and low to moderate density were found in harvested areas. Unharvested areas provided tall, dense vegetation structure that was especially attractive to tall-grass bird species, such as sedge wren (Cistothorus platensis) and Henslow's sparrow (Ammodramus henslowii). When considering wildlife habitat value in harvest management of switchgrass for biofuel, leaving some fields unharvested each year would be a good compromise, providing some habitat for a larger number of grassland bird species of management concern than if all fields were harvested annually. In areas where most idle grassland habitat present on the landscape is tallgrass, harvest of switchgrass for biofuel has the potential to increase the local diversity of grassland birds.     

Optimisation and techno-economic analysis of autonomous photovoltaic–wind hybrid energy systems in comparison to single photovoltaic and wind systems

A techno-economic analysis for autonomous small scale photovoltaic–wind hybrid energy systems is undertaken for optimisation purposes in the present paper. The answer to the question whether a hybrid photovoltaic–wind or a single photovoltaic or wind system is techno-economically better is also sought. Monthly analysis of 8 year long measured hourly weather data shows that solar and wind resources vary greatly from one month to the next. The monthly combinations of these resources lead to basically three types of months: solar-biased month, wind-biased month and even month. This, in turn, leads to energy systems in which the energy contributions from photovoltaic and wind generators vary greatly. The monthly and yearly system performances simulations for different types of months show that the system performances vary greatly for varying battery storage capacities and different fractions of photovoltaic and wind energy. As well as the system performance, the optimisation process of such hybrid systems should further consist of the system cost. Therefore, the system performance results are combined with system cost data. The total system cost and the unit cost of the produced electricity (for a 20 year system lifetime) are analysed with strict reference to the yearly system performance. It is shown that an optimum combination of the hybrid photovoltaic–wind energy system provides higher system performance than either of the single systems for the same system cost for every battery storage capacity analysed in the present study. It is also shown that the magnitude of the battery storage capacity has important bearings on the system performance of single photovoltaic and wind systems. The single photovoltaic system performs better than a single wind system for 2 day storage capacity, while the single wind system performs better for 1.25 day storage capacity for the same system cost.     

风能-太阳能互补独立发电系统设计优化与分析

提出了一种风能-太阳能互补发电系统的优化设计方法.在经济技术性能分析过程中引入了Pareto最优的概念来确定最优系统方案.实例计算表明:使用该优化设计方法可以获得系统经济技术性能与系统配置的关系和大量的Pareto最优解,从而可以得到最满意的系统方案.风能发电和太阳能发电具有互补性,风能-太阳能互补发电系统的经济技术性能优于单一的风能发电或太阳能发电系统.     

基于卷积和共享权系数长短时记忆网络时融合机制在风速预测上的应用研究

针对对于风能规划和应用都具有重大影响的风速存在强随机性问题,该文提出结合卷积神经网络（CNN）和共享权重长短期记忆网络（SWLSTM）的空时融合模型（CSWLSTM）,充分提取风速序列中蕴含的空域和时域信息,以提升预测精度。此外,为了获得可靠的风速概率预测结果,提出一种新的结合CNN、SWLSTM和高斯过程回归（GPR）的混合模型,称为 CSWLSTM-GPR。将CSWLSTM-GPR应用于中国内蒙古风速预测案例,从点预测精度、区间预测适用性和概率预测综合性能3个方面与相同结构的CNN和SWLSTM模型的风速预测方法进行比较。CSWLSTM-GPR的可靠性测试保证了预测结果的可靠性和说服力。实验结果表明,CSWLSTM-GPR在风速预测问题上能获得高精度的点预测、合适的预测区间和可靠的概率预测结果,也充分展现了该研究所提出CSWLSTM在风速预测方面具有较好的应用潜力。     

Wind turbine angular frequency analysis by means of computer vision techniques

The turbine synchronization phenomenon is of great interest in order to estimate the flicker produced by a wind farm. This paper proposes an initial approach to analyze the appearance of this phenomenon by the use of various image processing techniques: a method to automatically calculate the angular frequency of an unknown number of wind turbines from a video. The recorded video images were obtained at the Manzanal wind farm, province of León (Spain).     

基于ANSYS软件及正交试验方法的防覆冰摄像头结构参数最优化取值研究

为克服现有视频监控系统在冰期无法正常使用的困难,研究具备防冰、低功耗功能摄像机的最优结构尺寸具有重要意义。采用有限元软件ANSYS、流体力学Fluent计算模块,首先根据摄像机结构组成特征,设计9种正交试验设计模型,以加热玻璃表面的平均温度和水滴收集系数为优化目标,确定摄像机最优结构尺寸;其次根据确定的最优尺寸结构,研究外部风速、温度对摄像机内部的影响;最后根据露点温度计算方法,对防冰摄像机实际功耗进行测试。结果表明,摄像机尺寸最佳配比为H1∶H2∶H3=235∶200∶50,在外部环境风速为10m/s,最低温度为-10℃环境下腔体玻璃表面及摄像头均能达到无覆冰效果,同时在极端低温-20℃的环境下,一小时功耗仅为1.25W,远远满足实际应用情况。研究成果能够有效解决现有摄像头在冰期使用中存在的弊端,同时对保障输电线路冰期稳定运行、防冰管控具有重要意义。     

基于深度卷积自编码器的风电机组故障预警方法研究

通过风电机组状态监测进行故障预警,可防止故障进一步发展,降低风场运维成本。为充分挖掘风电机组监控与数据采集（SCADA）各状态参数时序信息,以及不同参数之间的非线性关系,该文将深度学习中自动编码器（AE）与卷积神经网络（CNN）相结合,提出基于深度卷积自编码（DCAE）的风电机组状态监测故障预警方法。首先基于历史SCADA数据离线建立基于DCAE的机组正常运行状态模型,然后分析重构误差确定告警阈值,使用EMWA控制图对实时对机组状态监测并进行故障预警。以北方某风电场2 MW双馈型风电机组叶片故障为实例进行实验分析,结果表明该文提出DCAE状态监测故障预警方法,可有效对机组故障提前预警,且优于现有基于深度学习的风电机组故障预警方法,可显著提升重构精度、减少模型参数和训练时间。     

Review of contribution to frequency control through variable speed wind turbine

The increasing wind penetration in today’s power grids has led to growing interest in the frequency control capabilities of wind generation. Several publications have proposed a variety of methods both on the levels of a single turbine and of a wind farm. This paper focuses on the role of wind generation in a system’s primary frequency control. Wind turbine control methods that enable frequency support and control are presented. The advantages and disadvantages of each method are discussed.     

高压共轨燃油喷射雾化特性试验研究

基于GS-1000型高压共轨试验台和定容容器,利用高速摄影技术对高压共轨燃油喷射的雾化特性进行了研究。通过计算机采集多种工况下的喷雾雾化图像,并利用C#语言编制的图像处理软件对喷雾图像进行了精确处理,测取了喷雾贯穿距、喷雾锥角等喷雾特性参数。试验结果表明：喷雾贯穿距随喷射压力的升高而增大,喷射压力对喷雾锥角的影响不明显。燃油喷雾贯穿距随背压的升高而减小,喷雾锥角随背压的升高而增大。     

Application analysis of a superconducting fault current limiter-magnetic energy storage system for the wind farm

风力发电所面临的两大重要问题是低电压穿越能力弱和功率输出不稳定。为了同时解决这两个问题,我们提出了超导限流-储能系统,并进行了单机系统的仿真研究,证实了该方案的有效性。然而对于风电场的应用,目前尚无研究。本文将超导限流-储能系统的应用扩展到风电场,分析了其提高低电压穿越能力和稳定有功功率输出的机理,并进行了仿真研究。从仿真结果来看,超导限流-储能系统能够同时提高风电场所有风机的低电压穿越能力,并能有效地平滑整个风电场的有功输出功率。考虑不同风机的互补效应,将该系统应用于风力发电场与直接应用于单台风机相比,其储能量和功率输出的要求可以大大降低,从而可以有效地减少系统总成本,因而具有更好的应用前景。     

Monitoring combustion systems using HMM probabilistic reasoning in dynamic flame images

In this paper, a novel method of on-line flame detection in video is proposed. Processing the data generated by an ordinary camera monitoring scene, it aims to early detect the current state of the combustion system and prevent the system from further degradation and occurrence of failure. Due to the dynamic change of the combustion system, the turbulent flame flicker produces images with different spatial and high temporal resolutions. The proposed method consists of hidden Markov model (HMM) and multiway principal component analysis (MPCA). MPCA is used to extract the cross-correlation among spatial relationships in the low dimensional space while HMM constructs the temporal behavior of the sequential observation. Although the prior process knowledge may not be available in the operation processes, the probability distribution of the normal status can be trained by the images collected from the normal operation processes. Subsequently, monitoring of a new observed image is achieved by a recursive Viterbi algorithm which can find the transition state sequence from series of observed image data. The proposed method, like the philosophy of traditional statistical process control, can generate simple probability monitoring charts to track the progress of the current transition state sequence and monitor the occurrence of the observable upsets. The advantages of the proposed method, data from the monitoring practice in the real combustion systems, are presented to help readers delve into the matter.