Transactions of Nanjing University of Aeronautics & Astronautics
Volume 36,Issue 6,2019 Table of Contents
2019, 36(6):879-888.
DOI: 10.16356/j.1005-1120.2019.06.001
Abstract:
Dual-frequency satellite positioning receivers are widely used because they can eliminate ionospheric delay and solve the full-circumference ambiguity quickly. However, in traditional dual-frequency receivers, the relevance of dual-frequency signals are not considered, and, with no improvement imposed to the tracking loop, two independent tracking loops are used to achieve the tracking of dual-frequency signals. In this paper, the BeiDou dual-frequency signals joint tracking algorithm based on Kalman filter is proposed for the tracking of BeiDou B1I and B3I dual-frequency signals. Taking the relevance of B1I and B3I signals into consideration, the algorithm adds a Kalman filter between the phase detector and carrier loop filter of the traditional dual-frequency independent tracking loop. The output results of the phase detectors of the B1I and B3I branches are then combined and filtered by the Kalman filter, and the results are input to the carrier loop filters of the corresponding branches. Proved by experiments, the algorithm not only enables the loop to enter a stable tracking state quickly, but also reduces the noise bandwidth of the two loop filters by about 10 Hz with the same tracking performance obtained.
Dual-frequency satellite positioning receivers are widely used because they can eliminate ionospheric delay and solve the full-circumference ambiguity quickly. However, in traditional dual-frequency receivers, the relevance of dual-frequency signals are not considered, and, with no improvement imposed to the tracking loop, two independent tracking loops are used to achieve the tracking of dual-frequency signals. In this paper, the BeiDou dual-frequency signals joint tracking algorithm based on Kalman filter is proposed for the tracking of BeiDou B1I and B3I dual-frequency signals. Taking the relevance of B1I and B3I signals into consideration, the algorithm adds a Kalman filter between the phase detector and carrier loop filter of the traditional dual-frequency independent tracking loop. The output results of the phase detectors of the B1I and B3I branches are then combined and filtered by the Kalman filter, and the results are input to the carrier loop filters of the corresponding branches. Proved by experiments, the algorithm not only enables the loop to enter a stable tracking state quickly, but also reduces the noise bandwidth of the two loop filters by about 10 Hz with the same tracking performance obtained.
2 Impact Analysis of Solar Irradiance Change on Precision Orbit Determination of Navigation Satellites
2019, 36(6):889-901.
DOI: 10.16356/j.1005-1120.2019.06.002
Abstract:
Solar radiation pressure is the main driving force and error source for precision orbit determination of navigation satellites. It is proportional to the solar irradiance, which is the “sun constant”. In regular calculation, the “solar constant” is regard as a constant. However, due to the existence of sunspots, flares, etc., the solar constant is not fixed, the change in the year is about 1%. To investigate the variation of solar irradiance, we use interpolation and average segment modeling of total solar irradiance data of SORCE, establishing variance solar radiation pressure (VARSRP) model and average solar radiation pressure (AVESRP) model based on the built solar pressure model (SRPM) (constant model). According to observation data of global positioning system (GPS) and Beidou system (BDS) in 2015 and comparing the solar pressure acceleration of VARSRP, AVESRP and SRPM, the magnitude of change can reach . In addition, according to the satellite precise orbit determination, for GPS satellites, the results of VARSRP and AVESRP are slightly smaller than those of the SRPM model, and the improvement is between 0.1 to 0.5 mm. For geosynchronous orbit (GEO) satellites of BDS, The AVESRP and VARSRP have an improvement of 3.5 mm and 4.0 mm, respectively, based on overlapping arc, and SLR check results show the AVESRP model and the VARSRP model is improved by 2.3 mm and 3.5 mm, respectively. Moreover, the change of inclined geosynchronous orbit (IGSO) satellites and medium earth orbit (MEO) satellites is relatively small, and the improvement is smaller than 0.5 mm.
Solar radiation pressure is the main driving force and error source for precision orbit determination of navigation satellites. It is proportional to the solar irradiance, which is the “sun constant”. In regular calculation, the “solar constant” is regard as a constant. However, due to the existence of sunspots, flares, etc., the solar constant is not fixed, the change in the year is about 1%. To investigate the variation of solar irradiance, we use interpolation and average segment modeling of total solar irradiance data of SORCE, establishing variance solar radiation pressure (VARSRP) model and average solar radiation pressure (AVESRP) model based on the built solar pressure model (SRPM) (constant model). According to observation data of global positioning system (GPS) and Beidou system (BDS) in 2015 and comparing the solar pressure acceleration of VARSRP, AVESRP and SRPM, the magnitude of change can reach
2019, 36(6):902-911.
DOI: 10.16356/j.1005-1120.2019.06.003
Abstract:
This paper presents a cascode configuration synchronous rectifier device based on silicon MOSFET and Schottky diode, which can replace traditional power diode directly. This structure has self-driven ability with simple external circuit, and the conduction characteristic is preferable to a power diode. Static characterization and switching behavior analysis of proposed structure are conducted in this paper. The switching process is illustrated in detail using real model which considers the parasitic inductances and the nonlinearity of junction capacitors. The real time internal voltage and current value during switching transition are deduced with the equivalent circuit. To validate the analysis, two voltage specification rectifiers are built. Finally, double-pulse test results and the practical design example verify the performance advantages of proposed structure.
This paper presents a cascode configuration synchronous rectifier device based on silicon MOSFET and Schottky diode, which can replace traditional power diode directly. This structure has self-driven ability with simple external circuit, and the conduction characteristic is preferable to a power diode. Static characterization and switching behavior analysis of proposed structure are conducted in this paper. The switching process is illustrated in detail using real model which considers the parasitic inductances and the nonlinearity of junction capacitors. The real time internal voltage and current value during switching transition are deduced with the equivalent circuit. To validate the analysis, two voltage specification rectifiers are built. Finally, double-pulse test results and the practical design example verify the performance advantages of proposed structure.
2019, 36(6):912-918.
DOI: 10.16356/j.1005-1120.2019.06.004
Abstract:
An ant colony optimization with artificial potential field (ACOAPF) algorithm is proposed to solve the cooperative search mission planning problem of unmanned aerial vehicle(UAV) swarm. This algorithm adopts a distributed architecture where each UAV is considered as an ant and makes decision autonomously. At each decision step, the ants choose the next gird according to the state transition rule and update its own artificial potential field and pheromone map based on the current search results. Through iterations of this process, the cooperative search of UAV swarm for mission area is realized. The state transition rule is divided into two types. If the artificial potential force is larger than a threshold, the deterministic transition rule is adopted, otherwise a heuristic transition rule is used. The deterministic transition rule can ensure UAVs to avoid the threat or approach the target quickly. And the heuristics transition rule considering the pheromone and heuristic information ensures the continuous search of area with the goal of covering more unknown area and finding more targets. Finally, simulations are carried out to verify the effectiveness of the proposed ACOAPF algorithm for cooperative search mission of UAV swarm.
An ant colony optimization with artificial potential field (ACOAPF) algorithm is proposed to solve the cooperative search mission planning problem of unmanned aerial vehicle(UAV) swarm. This algorithm adopts a distributed architecture where each UAV is considered as an ant and makes decision autonomously. At each decision step, the ants choose the next gird according to the state transition rule and update its own artificial potential field and pheromone map based on the current search results. Through iterations of this process, the cooperative search of UAV swarm for mission area is realized. The state transition rule is divided into two types. If the artificial potential force is larger than a threshold, the deterministic transition rule is adopted, otherwise a heuristic transition rule is used. The deterministic transition rule can ensure UAVs to avoid the threat or approach the target quickly. And the heuristics transition rule considering the pheromone and heuristic information ensures the continuous search of area with the goal of covering more unknown area and finding more targets. Finally, simulations are carried out to verify the effectiveness of the proposed ACOAPF algorithm for cooperative search mission of UAV swarm.
2019, 36(6):919-926.
DOI: 10.16356/j.1005-1120.2019.06.005
Abstract:
This paper focuses on the application of H ∞ preview control in automatic carrier landing system (ACLS) for carrier-based aircraft. Due to the mutual movement between aircraft and carrier, the landing process becomes considerably more challenging compared to a conventional runway landing. ACLS systems mitigate this by predicting deck motion and generating ideal glide slope path for tracking. Although, this predicted glide slope information is available in advance, conventional control structures are still unable to use this future information. H ∞ preview control has the ability to utilize this future information for improving tracking response and disturbance rejection. The process of incorporating preview information into ACLS framework and synthesizing the H ∞ preview controller is presented. The methodology is verified using the example of F/A-18 automatic carrier landing problem and results are presented.
This paper focuses on the application of H ∞ preview control in automatic carrier landing system (ACLS) for carrier-based aircraft. Due to the mutual movement between aircraft and carrier, the landing process becomes considerably more challenging compared to a conventional runway landing. ACLS systems mitigate this by predicting deck motion and generating ideal glide slope path for tracking. Although, this predicted glide slope information is available in advance, conventional control structures are still unable to use this future information. H ∞ preview control has the ability to utilize this future information for improving tracking response and disturbance rejection. The process of incorporating preview information into ACLS framework and synthesizing the H ∞ preview controller is presented. The methodology is verified using the example of F/A-18 automatic carrier landing problem and results are presented.
2019, 36(6):927-937.
DOI: 10.16356/j.1005-1120.2019.06.006
Abstract:
The electrolytic-capacitor-less (EL-cap-less) technique has been attracting more and more interests in these years. The EL-cap-less single-phase to three-phase inverter (STI) with front-end diode rectifier is applied to home electrical appliances. In EL-cap-less systems, film-type capacitor is used to suppress the high-frequency ripple components instead of smoothing the DC-link voltage. By using the EL-cap-less solution, reliability can be improved, lifetime can be increased and high power factor (PF) also can be realized. The EL-cap-less STI brings the opportunity to simplify system structure and lower cost. Because a certain DC-link capacitance is still required, the real PF is limited. To further improve the PF, the ripple components of current references are made in synchronous with the DC-link voltage. And the capacitance chosen guideline of the film-type capacitor also be presented. A prototype is built to verify its real performance.
The electrolytic-capacitor-less (EL-cap-less) technique has been attracting more and more interests in these years. The EL-cap-less single-phase to three-phase inverter (STI) with front-end diode rectifier is applied to home electrical appliances. In EL-cap-less systems, film-type capacitor is used to suppress the high-frequency ripple components instead of smoothing the DC-link voltage. By using the EL-cap-less solution, reliability can be improved, lifetime can be increased and high power factor (PF) also can be realized. The EL-cap-less STI brings the opportunity to simplify system structure and lower cost. Because a certain DC-link capacitance is still required, the real PF is limited. To further improve the PF, the ripple components of current references are made in synchronous with the DC-link voltage. And the capacitance chosen guideline of the film-type capacitor also be presented. A prototype is built to verify its real performance.
2019, 36(6):938-951.
DOI: 10.16356/j.1005-1120.2019.06.007
Abstract:
The aerodynamic model of propeller, wing, fuselage and vertical tail are established for the tilt quad rotor (TQR) with partial tilt-wing, and then the flight dynamic model is established. Based on the six-degree-of-freedom equation and the small disturbance linearization assumption, the trimming and stability of the tilt quad rotor with partial tilt-wing and the tilt quad rotor without tilt-wing are analyzed. The results show that in the hovering state, due to the existence of tilt-wing, the propeller wake reduces the downwash on the wing, thereby reducing the vertical weight gain of the aircraft. It is beneficial to increase the endurance time and improve the endurance performance. The transition corridor of the TQR with tilt-wing is narrower than that of the TQR without tilt-wing, but the transition corridor of TQR with tilt-wing still has a large space for design. Furthermore， the stability analysis shows that the Dutch roll damping ratio is larger, and in other modes the aircraft has a certain stability. The manipulation response analysis shows that in the transition mode the lateral-directional coupling is strong.
The aerodynamic model of propeller, wing, fuselage and vertical tail are established for the tilt quad rotor (TQR) with partial tilt-wing, and then the flight dynamic model is established. Based on the six-degree-of-freedom equation and the small disturbance linearization assumption, the trimming and stability of the tilt quad rotor with partial tilt-wing and the tilt quad rotor without tilt-wing are analyzed. The results show that in the hovering state, due to the existence of tilt-wing, the propeller wake reduces the downwash on the wing, thereby reducing the vertical weight gain of the aircraft. It is beneficial to increase the endurance time and improve the endurance performance. The transition corridor of the TQR with tilt-wing is narrower than that of the TQR without tilt-wing, but the transition corridor of TQR with tilt-wing still has a large space for design. Furthermore， the stability analysis shows that the Dutch roll damping ratio is larger, and in other modes the aircraft has a certain stability. The manipulation response analysis shows that in the transition mode the lateral-directional coupling is strong.
2019, 36(6):952-963.
DOI: 10.16356/j.1005-1120.2019.06.008
Abstract:
The low-stiffness of aircraft skins may results in the differences between aircraft actual parts and their theoretical models, which will consequently affect the accuracy of automatic drilling and riveting in aircraft assembly. In this paper, a novel approach of hole position correction using laser line scanner (LLS) is proposed to assign a single row of holes on the parts’ surfaces. First, we adopt a space circle fitting method and the random sample consensus (RANSAC) to obtain the precise coordinates of center of the datum holes’ coordinates. Second, LLS is calibrated by the laser tracker, and the relations between the LLS coordinate system and the tool coordinate system (TCS) can be calculated. Third, the kinematics model of the automatic riveting machine is established based on a two-point referencing strategy proposed in this paper. Thus, the positions of the holes to be drilled can be adjusted. Finally, the experimental results show that in TCS the measurement error of LLS is less than 0.1 mm, and the correction error of the hole position is less than 0.5 mm, which demonstrates the reliability of our method.
The low-stiffness of aircraft skins may results in the differences between aircraft actual parts and their theoretical models, which will consequently affect the accuracy of automatic drilling and riveting in aircraft assembly. In this paper, a novel approach of hole position correction using laser line scanner (LLS) is proposed to assign a single row of holes on the parts’ surfaces. First, we adopt a space circle fitting method and the random sample consensus (RANSAC) to obtain the precise coordinates of center of the datum holes’ coordinates. Second, LLS is calibrated by the laser tracker, and the relations between the LLS coordinate system and the tool coordinate system (TCS) can be calculated. Third, the kinematics model of the automatic riveting machine is established based on a two-point referencing strategy proposed in this paper. Thus, the positions of the holes to be drilled can be adjusted. Finally, the experimental results show that in TCS the measurement error of LLS is less than 0.1 mm, and the correction error of the hole position is less than 0.5 mm, which demonstrates the reliability of our method.
2019, 36(6):964-976.
DOI: 10.16356/j.1005-1120.2019.06.009
Abstract:
Components of mechanical product are assembled by structural joints, such as bolting, riveting, welding, etc. Structural joints introduce nonlinearity to some engineering structures, and the nonlinearity need to be modeled precisely. To meet serious quality requirements, it is necessary to detect and identify nonlinearity of mechanical products for structural optimization. Modal test to acquire a dynamic response has been applied for decades, which provides reliable results for finite element (FE) model updating. Here response control vibration test for identification of nonlinearity is presented. A nonlinear system can be regarded as linearity for particular steady state response, and classical linear analysis tool is applicable to extract modal data for particular response. First, its applicability is illustrated by some numerical simulations. Subsequently, it is implemented on experimental setup with structural joints by shaking table. The stiffness and damping function dependent of relative displacement are fitted to describe its inherent nonlinearity. The spring and damping forces are identified by harmonic balance method (HBM) to predict output response. Based on the identified results, the procedure is recommended that it allows a reliable measurement of nonlinearity with a certain accuracy.
Components of mechanical product are assembled by structural joints, such as bolting, riveting, welding, etc. Structural joints introduce nonlinearity to some engineering structures, and the nonlinearity need to be modeled precisely. To meet serious quality requirements, it is necessary to detect and identify nonlinearity of mechanical products for structural optimization. Modal test to acquire a dynamic response has been applied for decades, which provides reliable results for finite element (FE) model updating. Here response control vibration test for identification of nonlinearity is presented. A nonlinear system can be regarded as linearity for particular steady state response, and classical linear analysis tool is applicable to extract modal data for particular response. First, its applicability is illustrated by some numerical simulations. Subsequently, it is implemented on experimental setup with structural joints by shaking table. The stiffness and damping function dependent of relative displacement are fitted to describe its inherent nonlinearity. The spring and damping forces are identified by harmonic balance method (HBM) to predict output response. Based on the identified results, the procedure is recommended that it allows a reliable measurement of nonlinearity with a certain accuracy.
2019, 36(6):977-985.
DOI: 10.16356/j.1005-1120.2019.06.010
Abstract:
The injection characteristics of the main fuel nozzle, which is widely applied in advanced lean-premixed-prevaporized (LPP) low-emission combustors, can be simplified as the atomization and vaporization processes of a jet into cross-flow. In this study, a nozzle with a diameter of 0.4 mm is designed and processed through the heating of the inlet air, and the vaporization characteristics are investigated. The optical measurement and cyclone separation methods are separately used to investigate the evaporation rate of a jet into cross-flow. Experimental results show that the fuel evaporation rate in cross-flow is mainly affected by the Weber number (We), equivalent ratio (φ), momentum rate of fuel to air (q), and air temperature. In addition, the inlet temperature is a crucial factor for the evaporation ratio of a jet into cross-flow. The evaporation results measured by two different methods in the same cross-flow are very close to each other with a deviation within 10%.
The injection characteristics of the main fuel nozzle, which is widely applied in advanced lean-premixed-prevaporized (LPP) low-emission combustors, can be simplified as the atomization and vaporization processes of a jet into cross-flow. In this study, a nozzle with a diameter of 0.4 mm is designed and processed through the heating of the inlet air, and the vaporization characteristics are investigated. The optical measurement and cyclone separation methods are separately used to investigate the evaporation rate of a jet into cross-flow. Experimental results show that the fuel evaporation rate in cross-flow is mainly affected by the Weber number (We), equivalent ratio (φ), momentum rate of fuel to air (q), and air temperature. In addition, the inlet temperature is a crucial factor for the evaporation ratio of a jet into cross-flow. The evaporation results measured by two different methods in the same cross-flow are very close to each other with a deviation within 10%.
11 Modeling and Characteristic Analysis of Visco-Hyperelastic Film Based on Biaxial Tensile Experiment
2019, 36(6):986-994.
DOI: 10.16356/j.1005-1120.2019.06.011
Abstract:
Dielectric elastomers(DEs) show complex mechanical behaviors with different boundary conditions, geometry sizes, and prestress. In this study, a three-component linear visco-hyperelastic model of DE film is developed based on equibiaxial tension. By applying hereditary integrals to analyze multiple-segment loading processesof film stretching, the model parameters are extracted by fitting the visco-hyperelastic film model to the experimental data. To demonstrate the performance of proposed model, the obtained predictive results are compared with the experimental results under different equibiaxial loading conditions. The good agreement between them shows that the linear visco-hyperelastic model is a promising tool for analytically investigating the property of pre-stretched DE actuators. Finally, the Prony series coefficients are calculated according to the relaxation function. This is helpful for simulation analysis using commercial finite element software.
Dielectric elastomers(DEs) show complex mechanical behaviors with different boundary conditions, geometry sizes, and prestress. In this study, a three-component linear visco-hyperelastic model of DE film is developed based on equibiaxial tension. By applying hereditary integrals to analyze multiple-segment loading processesof film stretching, the model parameters are extracted by fitting the visco-hyperelastic film model to the experimental data. To demonstrate the performance of proposed model, the obtained predictive results are compared with the experimental results under different equibiaxial loading conditions. The good agreement between them shows that the linear visco-hyperelastic model is a promising tool for analytically investigating the property of pre-stretched DE actuators. Finally, the Prony series coefficients are calculated according to the relaxation function. This is helpful for simulation analysis using commercial finite element software.
2019, 36(6):995-1003.
DOI: 10.16356/j.1005-1120.2019.06.012
Abstract:
The capture operation performed by a snare-type end-effector mainly relies on three flexible cables. This paper solves the dynamics modeling problems of flexible cable used in the snare-type end-effector and provides a contact tracking control strategy for the impact phase of snare capture. To describe the motion of flexible cable, a dynamics model is established by considering both tensile and bending resistance properties. On this basis, a virtual spring concept is introduced to represent the contact between flexible cables and the target grapple shaft, and a contact dynamics model is established approximately by polynomial function with the variables of penetration and start-end distance of flexible cable. Thereafter, a contact tracking control strategy is proposed to improve the reliability of space snare capture. The target grapple shaft and flexible cable can keep in contact at the initial contact point during the whole capture process and thus reduce the possibility of pushing the target away. Experiments are carried out to verify the effectiveness of the proposed method.
The capture operation performed by a snare-type end-effector mainly relies on three flexible cables. This paper solves the dynamics modeling problems of flexible cable used in the snare-type end-effector and provides a contact tracking control strategy for the impact phase of snare capture. To describe the motion of flexible cable, a dynamics model is established by considering both tensile and bending resistance properties. On this basis, a virtual spring concept is introduced to represent the contact between flexible cables and the target grapple shaft, and a contact dynamics model is established approximately by polynomial function with the variables of penetration and start-end distance of flexible cable. Thereafter, a contact tracking control strategy is proposed to improve the reliability of space snare capture. The target grapple shaft and flexible cable can keep in contact at the initial contact point during the whole capture process and thus reduce the possibility of pushing the target away. Experiments are carried out to verify the effectiveness of the proposed method.
2019, 36(6):1004-1017.
DOI: 10.16356/j.1005-1120.2019.06.013
Abstract:
Based on the weak formulation for combined surface diffusion and evaporation-condensation, a governing equation of the finite element is derived for simulating the evolution of intergranular microcracks in copper interconnects induced simultaneously by stressmigration, electromigration and thermomigration. Unlike previously published works, the effect of thermomigration is considered. The results show that thermomigration can contribute to the microcrack splitting and accelerate the drifting process along the direction of the electric field. The evolution of the intergranular microcracks depends on the mechanical stress field, the temperature gradient field, the electric field, the initial aspect ratio and the linewidth. And there exists a critical electric field , a critical stress field , a critical aspect ratio and a critical linewidth . When , , or , the intergranular microcrack will split into two or three small intergranular microcracks. Otherwise, the microcrack will evolve into a stable shape as it migrates along the interconnect line. The critical stress field, the critical electric field and the critical aspect ratio decrease with a decrease in the linewidth, and the critical linewidth increases with an increase in the electric field and the aspect ratio. The increase of the stress field, the electric field or the aspect ratio and the decrease of the linewidth are not only beneficial for the intergranular microcrack to split but also accelerate the microcrack splitting process.
Based on the weak formulation for combined surface diffusion and evaporation-condensation, a governing equation of the finite element is derived for simulating the evolution of intergranular microcracks in copper interconnects induced simultaneously by stressmigration, electromigration and thermomigration. Unlike previously published works, the effect of thermomigration is considered. The results show that thermomigration can contribute to the microcrack splitting and accelerate the drifting process along the direction of the electric field. The evolution of the intergranular microcracks depends on the mechanical stress field, the temperature gradient field, the electric field, the initial aspect ratio and the linewidth. And there exists a critical electric field
2019, 36(6):1018-1025.
DOI: 10.16356/j.1005-1120.2019.06.014
Abstract:
Nested linear array enables to enhance localization resolution and achieve under-determined direction of arrival (DOA) estimation. In this paper, the traditional two-level nested linear array is improved to achieve more degrees of freedom (DOFs) and better angle estimation performance. Furthermore, a computationally efficient DOA estimation algorithm is proposed. The discrete Fourier transform (DFT) method is utilized to obtain coarse DOA estimates, and subsequently, fine DOA estimates are achieved by spatial smoothing multiple signals classification (SS-MUSIC) algorithm. Compared to SS-MUSIC algorithm, the proposed algorithm has the same estimation accuracy with lower computational complexity because the coarse DOA estimates enable to shrink the range of angle spectral search. In addition, the estimation of the number of signals is not required in advance by DFT method. Extensive simulation results testify the effectiveness of the proposed algorithm.
Nested linear array enables to enhance localization resolution and achieve under-determined direction of arrival (DOA) estimation. In this paper, the traditional two-level nested linear array is improved to achieve more degrees of freedom (DOFs) and better angle estimation performance. Furthermore, a computationally efficient DOA estimation algorithm is proposed. The discrete Fourier transform (DFT) method is utilized to obtain coarse DOA estimates, and subsequently, fine DOA estimates are achieved by spatial smoothing multiple signals classification (SS-MUSIC) algorithm. Compared to SS-MUSIC algorithm, the proposed algorithm has the same estimation accuracy with lower computational complexity because the coarse DOA estimates enable to shrink the range of angle spectral search. In addition, the estimation of the number of signals is not required in advance by DFT method. Extensive simulation results testify the effectiveness of the proposed algorithm.
2019, 36(6):1026-1038.
DOI: 10.16356/j.1005-1120.2019.06.015
Abstract:
With the development of data age, data quality has become one of the problems that people pay much attention to. As a field of data mining, outlier detection is related to the quality of data. The isolated forest algorithm is one of the more prominent numerical data outlier detection algorithms in recent years. In the process of constructing the isolation tree by the isolated forest algorithm, as the isolation tree is continuously generated, the difference of isolation trees will gradually decrease or even no difference, which will result in the waste of memory and reduced efficiency of outlier detection. And in the constructed isolation trees, some isolation trees cannot detect outlier. In this paper, an improved iForest-based method GA-iForest is proposed. This method optimizes the isolated forest by selecting some better isolation trees according to the detection accuracy and the difference of isolation trees, thereby reducing some duplicate, similar and poor detection isolation trees and improving the accuracy and stability of outlier detection. In the experiment, Ubuntu system and Spark platform are used to build the experiment environment. The outlier datasets provided by ODDS are used as test. According to indicators such as the accuracy, recall rate, ROC curves, AUC and execution time, the performance of the proposed method is evaluated. Experimental results show that the proposed method can not only improve the accuracy and stability of outlier detection, but also reduce the number of isolation trees by 20%—40% compared with the original iForest method.
With the development of data age, data quality has become one of the problems that people pay much attention to. As a field of data mining, outlier detection is related to the quality of data. The isolated forest algorithm is one of the more prominent numerical data outlier detection algorithms in recent years. In the process of constructing the isolation tree by the isolated forest algorithm, as the isolation tree is continuously generated, the difference of isolation trees will gradually decrease or even no difference, which will result in the waste of memory and reduced efficiency of outlier detection. And in the constructed isolation trees, some isolation trees cannot detect outlier. In this paper, an improved iForest-based method GA-iForest is proposed. This method optimizes the isolated forest by selecting some better isolation trees according to the detection accuracy and the difference of isolation trees, thereby reducing some duplicate, similar and poor detection isolation trees and improving the accuracy and stability of outlier detection. In the experiment, Ubuntu system and Spark platform are used to build the experiment environment. The outlier datasets provided by ODDS are used as test. According to indicators such as the accuracy, recall rate, ROC curves, AUC and execution time, the performance of the proposed method is evaluated. Experimental results show that the proposed method can not only improve the accuracy and stability of outlier detection, but also reduce the number of isolation trees by 20%—40% compared with the original iForest method.
Mobile website
Website Copyright © Transactions of Nanjing University of Aeronautics & Astronautics