Spatial relation extraction is the process of identifying geographic entities from text and determining their corresponding spatial relations.Traditional spatial relation extraction mainly uses rule-based pattern matc...Spatial relation extraction is the process of identifying geographic entities from text and determining their corresponding spatial relations.Traditional spatial relation extraction mainly uses rule-based pattern matching,supervised learning-based or unsupervised learning-based methods.However,these methods suffer from poor time-sensitive,high labor cost and high dependence on large-scale data.With the development of pre-trained language models greatly alleviating the shortcomings of traditional methods,supervised learning methods incorporating pre-trained language models have become the mainstream relation extraction methods.Pipeline extraction and joint extraction,as the two most dominant ideas of relation extraction,both have obtained good performance on different datasets,and whether to share the contextual information of entities and relations is the main differences between the two ideas.In this paper,we compare the performance of two ideas oriented to spatial relation extraction based on Chinese corpus data in the field of geography and verify which method based on pre-trained language models is more suitable for Chinese spatial relation extraction.We fine-tuned the hyperparameters of the two models to optimize the extraction accuracy before the comparison experiments.The results of the comparison experiments show that pipeline extraction performs better than joint extraction of spatial relation extraction for Chinese text data with sentence granularity,because different tasks have different focus on contextual information,and it is difficult to take account into the needs of both tasks by sharing contextual information.In addition,we further compare the performance of the two models with the rule-based template approach in extracting topological,directional and distance relations,summarize the shortcomings of this experiment and provide an outlook for future work.展开更多
In the design and troubleshooting of aero-engine pipeline,the vibration reduction of the pipeline system is often achieved by adjusting the hoop layout,provided that the shape of pipeline remains unchanged.However,in ...In the design and troubleshooting of aero-engine pipeline,the vibration reduction of the pipeline system is often achieved by adjusting the hoop layout,provided that the shape of pipeline remains unchanged.However,in reality,the pipeline system with the best antivibration performance may be obtained only by adjusting the pipeline shape.In this paper,a typical spatial pipeline is taken as the research object,the length of straight-line segment is taken as the design variable,and an innovative optimization method of avoiding vibration of aero-engine pipeline is proposed.The relationship between straight-line segment length and parameters that determine the geometric characteristics of the pipeline,such as the position of key reference points,bending angle,and hoop position,are derived in detail.Based on this,the parametric finite element model of the pipeline system is established.Taking the maximum first-order natural frequency of pipeline as the optimization objective and introducing process constraints and vibration avoidance constraints,the optimization model of the pipeline system is established.The genetic algorithm and the golden section algorithm are selected to solve the optimization model,and the relevant solution procedure is described in detail.Finally,two kinds of pipelines with different total lengths are selected to carry out a case study.Based on the analysis of the influence of straight-line segment length on the vibration characteristics of the pipeline system,the optimization methods developed in this paper are demonstrated.Results show that the developed optimization method can obtain the optimal single value or interval of the straight-line segment length while avoiding the excitation frequency.In addition,the optimization efficiency of the golden section algorithm is remarkably higher than that of the genetic algorithm for length optimization of a single straight-line segment.展开更多
Crude oil transportation through pipelines presents danger to communities along its path. In the Niger Delta region of Nigeria for instance, pipeline vandalism occurs indiscriminately and regularly, such that every se...Crude oil transportation through pipelines presents danger to communities along its path. In the Niger Delta region of Nigeria for instance, pipeline vandalism occurs indiscriminately and regularly, such that every segment of a pipeline network becomes a potential target and possibly source of oil spill hazard. In terms of pipeline hazard and risk distribution, the oil plume’s ability to migrate freely in wetlands and encroachment on pipeline right of ways by people increases chances of wider contact and exposure opportunities to inhabitants and the environment. Despite several efforts to mitigate pipeline hazards in the oil and gas sector, none has been effective in Nigeria partly due to paucity of data in public domain and poor public participation. Therefore considering the environmental and human health challenges associated with oil spills, an alternative method was developed using multi-criteria decision analysis to model 1) pipeline hazard zones, 2) potential pipeline impact radius, and 3) high consequence areas with four attribute layers, i.e. land cover, population, river and pipeline to encourage public participation. The model identified land use areas, communities and rivers likely to be susceptible to pipeline hazards and areas requiring regular monitoring and possible intervention. Meanwhile the model sensitivity test indicated that the river layer was most sensitive, while transferability was limited to similar criteria variables. The model can stimulate public participation in pipeline hazard management while policy makers and regulators would find it relevant in oil spill impact mitigation.展开更多
基金supported by the National Key Research and Development Program of China under[Grant number 2021YFB3900903]the National Natural Science Foundation of China under[Grant number 41971337].
文摘Spatial relation extraction is the process of identifying geographic entities from text and determining their corresponding spatial relations.Traditional spatial relation extraction mainly uses rule-based pattern matching,supervised learning-based or unsupervised learning-based methods.However,these methods suffer from poor time-sensitive,high labor cost and high dependence on large-scale data.With the development of pre-trained language models greatly alleviating the shortcomings of traditional methods,supervised learning methods incorporating pre-trained language models have become the mainstream relation extraction methods.Pipeline extraction and joint extraction,as the two most dominant ideas of relation extraction,both have obtained good performance on different datasets,and whether to share the contextual information of entities and relations is the main differences between the two ideas.In this paper,we compare the performance of two ideas oriented to spatial relation extraction based on Chinese corpus data in the field of geography and verify which method based on pre-trained language models is more suitable for Chinese spatial relation extraction.We fine-tuned the hyperparameters of the two models to optimize the extraction accuracy before the comparison experiments.The results of the comparison experiments show that pipeline extraction performs better than joint extraction of spatial relation extraction for Chinese text data with sentence granularity,because different tasks have different focus on contextual information,and it is difficult to take account into the needs of both tasks by sharing contextual information.In addition,we further compare the performance of the two models with the rule-based template approach in extracting topological,directional and distance relations,summarize the shortcomings of this experiment and provide an outlook for future work.
基金This work was supported by the Major Projects of Aero-Engines and Gas Turbines(J2019-I-0008-0008)the Fundamental Research Funds for the Central Universities of China(Grant No.N180312012).
文摘In the design and troubleshooting of aero-engine pipeline,the vibration reduction of the pipeline system is often achieved by adjusting the hoop layout,provided that the shape of pipeline remains unchanged.However,in reality,the pipeline system with the best antivibration performance may be obtained only by adjusting the pipeline shape.In this paper,a typical spatial pipeline is taken as the research object,the length of straight-line segment is taken as the design variable,and an innovative optimization method of avoiding vibration of aero-engine pipeline is proposed.The relationship between straight-line segment length and parameters that determine the geometric characteristics of the pipeline,such as the position of key reference points,bending angle,and hoop position,are derived in detail.Based on this,the parametric finite element model of the pipeline system is established.Taking the maximum first-order natural frequency of pipeline as the optimization objective and introducing process constraints and vibration avoidance constraints,the optimization model of the pipeline system is established.The genetic algorithm and the golden section algorithm are selected to solve the optimization model,and the relevant solution procedure is described in detail.Finally,two kinds of pipelines with different total lengths are selected to carry out a case study.Based on the analysis of the influence of straight-line segment length on the vibration characteristics of the pipeline system,the optimization methods developed in this paper are demonstrated.Results show that the developed optimization method can obtain the optimal single value or interval of the straight-line segment length while avoiding the excitation frequency.In addition,the optimization efficiency of the golden section algorithm is remarkably higher than that of the genetic algorithm for length optimization of a single straight-line segment.
文摘Crude oil transportation through pipelines presents danger to communities along its path. In the Niger Delta region of Nigeria for instance, pipeline vandalism occurs indiscriminately and regularly, such that every segment of a pipeline network becomes a potential target and possibly source of oil spill hazard. In terms of pipeline hazard and risk distribution, the oil plume’s ability to migrate freely in wetlands and encroachment on pipeline right of ways by people increases chances of wider contact and exposure opportunities to inhabitants and the environment. Despite several efforts to mitigate pipeline hazards in the oil and gas sector, none has been effective in Nigeria partly due to paucity of data in public domain and poor public participation. Therefore considering the environmental and human health challenges associated with oil spills, an alternative method was developed using multi-criteria decision analysis to model 1) pipeline hazard zones, 2) potential pipeline impact radius, and 3) high consequence areas with four attribute layers, i.e. land cover, population, river and pipeline to encourage public participation. The model identified land use areas, communities and rivers likely to be susceptible to pipeline hazards and areas requiring regular monitoring and possible intervention. Meanwhile the model sensitivity test indicated that the river layer was most sensitive, while transferability was limited to similar criteria variables. The model can stimulate public participation in pipeline hazard management while policy makers and regulators would find it relevant in oil spill impact mitigation.
