Average Distance Method Research Articles

An interpretable spacecraft flywheel system health status assessment method under perturbation

Abstract Health status assessment is an important measure for maintaining the safety of spacecraft flywheel systems. The influence of noise, sensor quality, and other disturbance factors can lead to a decrease in the reliability of the collected information. This can affect the model accuracy. Moreover, a loss of belief in the model is frequently caused by the opaque nature of the procedure and the incomprehensibility of the outcomes, particularly in fields such as aerospace. It is urgent to maintain the interpretability of the model and successfully identify the unreliability of the observed data. Therefore, this paper proposes a spacecraft flywheel system health status assessment method under perturbation based on interpretable belief rule base with attribute reliability (IBRB-r). First, the attribute reliability is calculated based on the average distance method, and a new fusion method of attribute reliability is proposed to reduce the interference of unreliable information. Then, a new interpretable constraint strategy is proposed to improve the rationality and interpretability of the parameters. Finally, the proposed method is validated by a case study of the health status assessment of a spacecraft flywheel system. Experiments show that the IBRB-r maintains high accuracy and interpretability under unreliable observation data.

A new health prediction model for a sensor network based on belief rule base with attribute reliability

Health prediction plays an essential role in improving the reliability of a sensor network by guiding the network maintenance. However, affected by interference factors in the real operational environment, the reliability of the monitoring information about the sensor network tends to decline, which affects the health prediction accuracy. Furthermore, the lack of monitoring information and high complexity of the network increase the difficulty of health prediction. To solve these three problems, this paper proposes a new sensor network health prediction model based on the belief rule base model with attribute reliability (BRB-r). The BRB-r model is an expert system that fully considers the qualitative knowledge and quantitative data of the sensor network. In addition, it can address the fuzziness and nondeterminacy of this qualitative knowledge. In the new model, the unreliable monitoring information of the sensor network is handled by the attribute reliability mechanism. The reliability of the sensor is calculated by the average distance method. Due to the effect of the fuzziness and nondeterminacy of expert knowledge, the health status of the sensor network cannot be accurately estimated by the initial health prediction model. Consequently, the optimization model for the health prediction model is established. Finally, a case study regarding a sensor network for oil storage tanks is conducted, and the validity of this method is demonstrated.

Point cloud simplification algorithm based on particle swarm optimization for online measurement of stored bulk grain

The simplification of 3D laser scanning point cloud is an important step of surface reconstruction and volume estimation of bulk grain in granary. This study presented an adaptive simplification algorithm based on particle swarm optimization (PSO). It introduced PSO into the average distance method, a conventional simplification method. The basic idea of this algorithm was to adaptively determine the optimal point reducing intervals of scanning lines according to original point cloud density by PSO. By using the 3D point cloud scanned from bulk grain surface in granary, the proposed algorithm was validated. Compared with the average distance method, the proposed algorithm obtained more evenly distributed point set, smaller reduction ratio (6.96%) and higher volume estimation accuracy (relative error was less than 3‰). The 3D laser scanner (GSLS003, Jilin University and SkyViTech Co., Ltd., Hangzhou, China) used in this study could scan the complete picture of the grain surface in a granary in one time, so the acquired point cloud data do not have to be jointed. For the good simplification performance and capability of updating the reducing interval at any moment, the proposed algorithm and the 3D laser scanner could be used to realize online real-time measurement of stored bulk grain volume in granary. Keywords: point cloud, simplification algorithm, particle swarm optimization (PSO), 3D laser scanning, large object, stored grain DOI: 10.3965/j.ijabe.20160901.1805 Citation: Shao Q, Xu T, Yoshino T, Zhao Y, Yang W, Zhu H. Point cloud simplification algorithm based on particle swarm optimization for online measurement of stored bulk grain. Int J Agric & Biol Eng, 2016; 9(1): 71－78.

Identification of tiller age categories based on morphogenetic responses of continuously stocked marandu palisade grass fertilised with nitrogen

Nitrogen fertilization augments tiller appearance and death, modifying the age profile of tiller population. Tillers from different age categories present distinct responses to environmental factors and management practices characterized by variation in their morphogenetic responses, which interfere in the dynamics of the herbage accumulation process. In that context, rates of leaf appearance and elongation and leaf lifespan were used as indicators to identify tiller age categories of continuously stocked marandu palisade grass managed at 30cm and subjected to nitrogen fertilization regimes using cluster analysis. Analysis was based on the average distance method using combined means of the response-variables studied, and the results used to build a dendogram to illustrate the arrangement of the resulting clusters. Three clusters were identified corresponding to the tiller age categories denominated young (tillers with age between 0 and 60 days), mature (between 60 and 120 days) and old (over 120 days). Discrimination of the tiller age categories according to nitrogen fertilization regime followed the same pattern of clustering. The fact demonstrates consistency and indicates the three age categories as reference for detailed studies regarding the ecophysiology of marandu palisade grass.

Cluster analysis of Dairy Herd Improvement data to discover trends in performance characteristics in large Upper Midwest dairy herds

Principal component analysis (PCA) is a variable reduction method used on over-parameterized data sets with a vast number of variables and a limited number of observations, such as Dairy Herd Improvement (DHI) data, to select subsets of variables that describe the largest amount of variance. Cluster analysis (CA) segregates objects, in this case dairy herds, into groups based upon similarity in multiple characteristics simultaneously. This project aimed to apply PCA to discover the subset of most meaningful DHI variables and to discover groupings of dairy herds with similar performance characteristics. Year 2011 DHI data was obtained for 557 Upper Midwest herds with test-day mean ≥200 cows (assumed mostly freestall housed), that remained on test for the entire year. The PCA reduced an initial list of 22 variables to 16. The average distance method of CA grouped farms based on best goodness of fit determined by the minimum cophenetic distance. Six groupings provided the optimal fitting number of clusters. Descriptive statistics for the 16 variables were computed per group. On observations of means, groups 1, 2, and 6 demonstrated the best performances in most variables, including energy-corrected milk, linear somatic cell score (log of somatic cell count), dry period intramammary infection cure rate, new intramammary infection risk, risk of subclinical intramammary infection at first test, age at first calving, days in milk, and Transition Cow Index. Groups 3, 4, and 5 demonstrated the worst mean performances in most the PCA-selected variables, including DIM, age at first calving, risk of subclinical intramammary infection at first test, and dry period intramammary infection cure rate. Groups 4 and 5 also had the worst mean herd performances in energy-corrected milk, Transition Cow Index, linear somatic cell score, and new intramammary infection risk. Further investigation will be conducted to reveal patterns of management associated with herd categorization. The PCA and CA should be used when describing the multivariate performance of dairy herds and whenever working with over-parameterized data sets, such as DHI databases.

Genetic dissimilarity and selection of putative mutants of Terra Maranhão plantain cultivar using the Ward-MLM strategy.

The objective of this study was to evaluate genetic variability and select putative mutants of Terra Maranhão plantain cultivar (AAB genome) subjected to gamma radiation based on agronomic data and inter simple sequence repeat molecular marker profiles using the Ward-MLM strategy. A total of 233 irradiated plants and 41 controls were assessed. The agronomic and molecular data were subjected to the Ward-MLM statistical algorithm in the SAS program. Cluster analysis was performed by the average distance method (UPGMA), based on the distance matrix of the Gower algorithm, and the cophenetic correlation coefficient calculated using the R software. The distance between the putative mutants ranged from 0.321 to 0.524, with an average distance of 0.426, and a cophenetic correlation coefficient of 0.79. Three putative mutants, which were selected based on the best agronomic traits and low height, will undergo further evaluation in the next stages of the banana breeding program at Embrapa. These results describe the first attempt of using combined data of Terra Maranhão plantain cultivar for the purpose of selecting mutants and assessing genetic variability.

Identification of intestinal pacemaker frequency through time-frequency ridge analysis of surface EEnG.

The non-invasive monitoring of the frequency of intestinal pacemaker activity (slow wave, SW) has an important diagnostic value. However the presence of noise, physiological interferences and spurious peaks of the spectral estimators can yield to misidentification of SW frequency when using conventional dominant frequency detection method. In this paper, two methods of ridge extraction from the time-frequency distribution of human surface electroenterogram (EEnG) are proposed for the identification and tracking of SW frequency in 13 recording sessions of 120 minutes in 13 healthy volunteers. The minimum average distance method, that includes of information of previous and subsequent windows of analysis, yields the best results in terms of providing ridges that are longer, with less interruptions and with more stable frequency values which better suit the ubiquity and rhythmicity characteristics of the intestinal SW. This technique permits to reduce misinterpretations of intestinal SW frequency which can be of great importance in diagnostic applications of EEnG.

Similar structures to the E-to-H helix unit in the globin-like fold are found in other helical folds.

A protein in the globin-like fold contains six alpha-helices, A, B, E, F, G and H. Among them, the E-to-H helix unit (E, F, G and H helices) forms a compact structure. In this study, we searched similar structures to the E-to-H helix of leghomoglobin in the whole protein structure space using the Dali program. Several similar structures were found in other helical folds, such as KaiA/RbsU domain and Type III secretion system domain. These observations suggest that the E-to-H helix unit may be a common subunit in the whole protein 3D structure space. In addition, the common conserved hydrophobic residues were found among the similar structures to the E-to-H helix unit. Hydrophobic interactions between the conserved residues may stabilize the 3D structures of the unit. We also predicted the possible compact regions of the units using the average distance method.

Evaluation of the Selection of the Initial Seeds for K-Means Algorithm

Clustering method is divided into hierarchical clustering, partitioning clustering, and more. K-Means algorithm is one of partitioning clustering methods and is adequate to cluster a lot of data rapidly and easily. The problem is it is too dependent on initial centers of clusters and needs the time of allocation and recalculation. We compare random method, max average distance method and triangle height method for selecting initial seeds in KMeans algorithm. It reduces total clustering time by minimizing the number of allocation and recalculation.

Scale adjustments for classifiers in high-dimensional, low sample size settings

Distance-based classifiers are generally considered to be effective at discriminating between populations that differ in location. Indeed, nearest-neighbour methods and the support vector machine are frequently used in very high-dimensional problems involving gene expression data, where it is believed that elevated levels of expression convey much of the information for classification. However, one problem inherent to distance-based classifiers is that scale differences can mask location differences. In consequence, such classifiers can have poor performance if the information for classification accumulates through a large number of relatively small location differences in data components, rather than via large differences. In this paper, we show that a simple adjustment for scale, applicable to a variety of distance-based classifiers, can remedy the problem. For some classifiers, such as those based on the support vector machine or the centroid method, scale corrections are important primarily in the case of small training-sample sizes. However, for other classifiers, including those based on nearest-neighbour and average-distance methods, scale adjustments are helpful more generally.

REGIONAL VULNERABILITY OF SNOW WATER RESOURCE IN CLIMATE CHANGE

Snow water equivalent (SWE) distribution in time series was acquired from SWE model with AMeDAS data, remote sensing data, and elevation data. The SWE model was composed of the snowfall model and the snowmelt model. The snowfall distribution was presumed using the weighted average distance method with AMeDAS data in the snowfall model. The snowmelt was estimated using the degree-day method in the snowmelt model. The snowmelt rate factor was decided by comparing the snow map with SWE distribution. To discuss the regional vulnerability of snow water resource, difference between SWEs of the little snow year and the heavy one was estimated. As results, the difference was 1000mm and over in Yamagata, Niigata, Toyama and its surroundings, which were considered as vulnerable region of SWE in Japan.

A new method to compare planned and achieved position of an orthopaedic implant

The present study describes an automatic method to evaluate the efficacy of a computer aided orthopaedic surgery system by comparing the position of the joint implant, as derived from post-operative computed tomography (CT) scans, to that planned by the surgeon before the operation. The method relies on two spatial registrations, one to align the post-operative femur with the pre-operative femur, the second to compute the planned versus achieved (PVA) accuracy as the roto-translation that registers the pre-operative implant position with the post-operative position. Two surface registration algorithms (a generic average distance minimisation and the specialised iterative closest point (ICP) method) were comparatively evaluated first on a set of test cases to measure the absolute accuracy and robustness with respect to peculiar situations such as a distant starting point. The average distance method failed the registration of one test case and showed peak errors of 0.97° on the rotations and 3.09 mm on the translations. The ICP method was found much more efficient and was able to register all test cases. The peak error was 0.44° on the rotations and 0.67 mm on the translations. The ICP method was then used to compute the PVA accuracy on six clinical cases treated with a CT-based planning system in combination with conventional surgical procedures. The method successfully processed all cases demonstrating the efficacy of the proposed procedure in the specific application.

A class of tests for detecting 'general' and 'focused' clustering of rare diseases.

This paper proposes a class of tests applicable to the detection of two types of disease clustering 'focused' and 'general' clustering. The former assesses the clustering of observed cases around the fixed point and the latter does not have any prior information on the centre of clustering. The proposed test for 'general' clustering is a generalization of the index for temporal clustering proposed by Tango in that it adjusts for differences in population densities and also in population distributions among categories of the counfounders such as age and sex. Simulation study shows that the proposed 'general' test outperformed the average distance method of Whittemore et al. in most of the cluster models considered.