Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8908875
Shan Ding, F. Long, Huijin Fan, Lei Liu, Yongji Wang
Obstacle detection is an important issue in the study of an unmanned airship, which helps the airship to avoid obstacles and reduces the risk of accidences. This paper establishes an obstacle detection network, which is obtained by inserting wisely some $1times 1$ and $3times 3$ convolutional layers at the beginning and the end of the YOLOv3-tiny network. The experimental results show that our novel network leads to a higher accuracy compared to YOLOv3-tiny while with a satisfied processing speed.
{"title":"A Novel YOLOv3-tiny Network for Unmanned Airship Obstacle Detection","authors":"Shan Ding, F. Long, Huijin Fan, Lei Liu, Yongji Wang","doi":"10.1109/DDCLS.2019.8908875","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908875","url":null,"abstract":"Obstacle detection is an important issue in the study of an unmanned airship, which helps the airship to avoid obstacles and reduces the risk of accidences. This paper establishes an obstacle detection network, which is obtained by inserting wisely some $1times 1$ and $3times 3$ convolutional layers at the beginning and the end of the YOLOv3-tiny network. The experimental results show that our novel network leads to a higher accuracy compared to YOLOv3-tiny while with a satisfied processing speed.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"17 1","pages":"277-281"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85277435","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
For modern chemical process often has multiple operational modes and monitoring data in the process are often nonlinear and high dimensional, a new fault detection method based on improved local tangent space alignment (LTSA) is proposed in this paper. Firstly, aiming at the characteristics of multi-modality of the chemical process, an improved LTSA algorithm is proposed in the paper, called correlation tangent space arrangement (CLTSA). In CLTSA, the variable $r$ is constructed and used to describe the relationship between the multivariate variables and reconstruct the global coordinates of monitoring data. Then, the incremental learning mechanism is introduced in CLTSA. For newly collected data, only some elements of the transition matrix need to be updated. And the matrix similarity statistics is established to maintain the size of the transition matrix, which has improved the efficiency of the algorithm. Finally, nonlinear principal elements in monitoring data are extracted through CLTSA and statistics $T^{2}$ and $SPE$ are used to monitor the change of the principal elements. When the monitored amount exceeds the threshold, it is determined that a fault has occurred in the chemical process. The simulation results of TE process show that the method proposed in the paper has a high fault detection rate and provides a new way for fault detection of multi-modal nonlinear chemical processes.
{"title":"Fault Detection of Multimodal Chemical Process Based on CLTSA","authors":"Yankun Han, Ruirui Huang, Yandong Hou, Qianshuai Cheng","doi":"10.1109/DDCLS.2019.8908884","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908884","url":null,"abstract":"For modern chemical process often has multiple operational modes and monitoring data in the process are often nonlinear and high dimensional, a new fault detection method based on improved local tangent space alignment (LTSA) is proposed in this paper. Firstly, aiming at the characteristics of multi-modality of the chemical process, an improved LTSA algorithm is proposed in the paper, called correlation tangent space arrangement (CLTSA). In CLTSA, the variable $r$ is constructed and used to describe the relationship between the multivariate variables and reconstruct the global coordinates of monitoring data. Then, the incremental learning mechanism is introduced in CLTSA. For newly collected data, only some elements of the transition matrix need to be updated. And the matrix similarity statistics is established to maintain the size of the transition matrix, which has improved the efficiency of the algorithm. Finally, nonlinear principal elements in monitoring data are extracted through CLTSA and statistics $T^{2}$ and $SPE$ are used to monitor the change of the principal elements. When the monitored amount exceeds the threshold, it is determined that a fault has occurred in the chemical process. The simulation results of TE process show that the method proposed in the paper has a high fault detection rate and provides a new way for fault detection of multi-modal nonlinear chemical processes.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"8 1","pages":"598-603"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83641733","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8908927
Mingxuan Sun, Xing Li
This paper is concerned with the convergence rate improvement of consensus of multi-agent systems, for which we introduce a generic attracting law (GAL), involving three terms which specify a generic action for improvement of convergence performance. The conventional double power-rate attracting law is modified for forming GAL, by adding a proportional term, and the convergence rate of the system can be dramatically improved. Through the two-phase analysis, an estimate for the upper bound of the settling time function is given, by which the obtained upper bound depends upon the initial state, and is finite without regard to the value of the initial state. The GAL is adopted for the purpose of consensus of multi-agent systems. A nonlinear protocol is designed to make the system undertaken achieve finite-duration consensus, and numerical results are presented to validate its effectiveness.
{"title":"Finite-Duration Consensus of Multi-Agent Systems Using a Generic Attracting Law","authors":"Mingxuan Sun, Xing Li","doi":"10.1109/DDCLS.2019.8908927","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908927","url":null,"abstract":"This paper is concerned with the convergence rate improvement of consensus of multi-agent systems, for which we introduce a generic attracting law (GAL), involving three terms which specify a generic action for improvement of convergence performance. The conventional double power-rate attracting law is modified for forming GAL, by adding a proportional term, and the convergence rate of the system can be dramatically improved. Through the two-phase analysis, an estimate for the upper bound of the settling time function is given, by which the obtained upper bound depends upon the initial state, and is finite without regard to the value of the initial state. The GAL is adopted for the purpose of consensus of multi-agent systems. A nonlinear protocol is designed to make the system undertaken achieve finite-duration consensus, and numerical results are presented to validate its effectiveness.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"37 1","pages":"691-696"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85932134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8908964
H. Lu, Renren Wang
This design mainly analyzes and designs a 3D printing equipment model based on extrusion curing of raw materials. The working principle of the device is that the CAD model of the building to be printed is first used, and the program is written directly by point-by-point comparison interpolation method. The program is then stored as a device control system. When printing is required, only the corresponding program is called. And the system will control the servo driver, drive the servo motor, the servo motor drives the mechanical arm sprinkler head to move on the XYZ axis through the gear drive. At the same time, the system sends out a signal to drive the feeding motor to drive the screw rod to start feeding. The electromagnetic valve of the nozzle opens and starts to print and construct the building according to the program written by interpolation algorithm.
{"title":"Model Analysis and Design of 3D Printing Control System Based on Cement Components","authors":"H. Lu, Renren Wang","doi":"10.1109/DDCLS.2019.8908964","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908964","url":null,"abstract":"This design mainly analyzes and designs a 3D printing equipment model based on extrusion curing of raw materials. The working principle of the device is that the CAD model of the building to be printed is first used, and the program is written directly by point-by-point comparison interpolation method. The program is then stored as a device control system. When printing is required, only the corresponding program is called. And the system will control the servo driver, drive the servo motor, the servo motor drives the mechanical arm sprinkler head to move on the XYZ axis through the gear drive. At the same time, the system sends out a signal to drive the feeding motor to drive the screw rod to start feeding. The electromagnetic valve of the nozzle opens and starts to print and construct the building according to the program written by interpolation algorithm.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"51 1","pages":"1373-1376"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85852036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8908959
Xiaoqian Luo, Xinxin Wang, Tai-Fang Li
The paper mainly studies the event-triggered control of switched linear systems with partially random coefficient using average dwell time switching technique. Samplings are active when a triggering condition is satisfied. On every sampling instant, new state and the switching information are transferred to controller. Compared with the zero-order holder, a dynamic controller is introduced in closed-loop to approximate the plant behavior. A sufficient condition guaranteeing stability is established by using Lyapunov function method. A related example is presented finally.
{"title":"Feedback Control of Switched Linear Systems with Event-Triggered Link","authors":"Xiaoqian Luo, Xinxin Wang, Tai-Fang Li","doi":"10.1109/DDCLS.2019.8908959","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908959","url":null,"abstract":"The paper mainly studies the event-triggered control of switched linear systems with partially random coefficient using average dwell time switching technique. Samplings are active when a triggering condition is satisfied. On every sampling instant, new state and the switching information are transferred to controller. Compared with the zero-order holder, a dynamic controller is introduced in closed-loop to approximate the plant behavior. A sufficient condition guaranteeing stability is established by using Lyapunov function method. A related example is presented finally.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"368 1","pages":"158-162"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86807465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8909044
Lang Liu, Weidong Yang, Hong Zhang, Huijin Fan, Bo Tao
In industrial process, fault isolation technology can identify the major variables leading to faults. Contribution plots and reconstruction-based methods are common tools, but both of them have the smearing effect. To solve this problem, Bayesian theory based method have been developed. Unfortunately, they usually have a high misdiagnosis rate when handling multiple faults with small magnitudes. In this paper, a new fault isolation method based on Bayesian information criterion is proposed. Firstly, the fault isolation problem is transformed into a mixed integer nonlinear programming problem. Then, to reduce the difficulty of calculation, the original problem is simplified into a series of nested mixed integer quadratic programming problems by using forward selection strategy. Finally, these problems can be solved by branch and bound algorithm. The effectiveness of the proposed method is verified by Monte Carlo simulation.
{"title":"A Variable Selection Method for Fault Isolation through Bayesian Information Criterion","authors":"Lang Liu, Weidong Yang, Hong Zhang, Huijin Fan, Bo Tao","doi":"10.1109/DDCLS.2019.8909044","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909044","url":null,"abstract":"In industrial process, fault isolation technology can identify the major variables leading to faults. Contribution plots and reconstruction-based methods are common tools, but both of them have the smearing effect. To solve this problem, Bayesian theory based method have been developed. Unfortunately, they usually have a high misdiagnosis rate when handling multiple faults with small magnitudes. In this paper, a new fault isolation method based on Bayesian information criterion is proposed. Firstly, the fault isolation problem is transformed into a mixed integer nonlinear programming problem. Then, to reduce the difficulty of calculation, the original problem is simplified into a series of nested mixed integer quadratic programming problems by using forward selection strategy. Finally, these problems can be solved by branch and bound algorithm. The effectiveness of the proposed method is verified by Monte Carlo simulation.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"41 1","pages":"1071-1076"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86264901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8908916
An Liu, Xiao-fei Lu, Shaolin Hu, Weihui He
The classical Hazard Rate Function (HRF) is typically used to make preventative maintenance (PM) decisions and always estimates HRF based on time to failure data. Even so, PM determination using conventional HRF is not suitable for systems with condition monitoring. The system is stopped checking and maintaining when the measured health state exceeds a predetermined threshold, which is always referred to as a soft failure. Since the application of HRF for PM decision making has great advantages, the classic HRF should be extended to combine soft and hard failure of systems with condition monitoring. In this paper, we define the HRFs of hard and soft failure for system under condition monitoring and propose a method to estimate the HRFs with data of failure time. We discuss in detail the relationship between these two HRFs and the classical HRF. With double stochastic processes (processes of degradation and measured healthy status), the properties of these HRFs are also researched. Further the optimal maintenance decisions are made for non-repairable and repairable systems upon these two types of HRFs. Eventually, the idea of this paper is verified by numerical examples.
{"title":"Health Maintenance Decisions Based on Hazard Rate Function under Degradation Process","authors":"An Liu, Xiao-fei Lu, Shaolin Hu, Weihui He","doi":"10.1109/DDCLS.2019.8908916","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908916","url":null,"abstract":"The classical Hazard Rate Function (HRF) is typically used to make preventative maintenance (PM) decisions and always estimates HRF based on time to failure data. Even so, PM determination using conventional HRF is not suitable for systems with condition monitoring. The system is stopped checking and maintaining when the measured health state exceeds a predetermined threshold, which is always referred to as a soft failure. Since the application of HRF for PM decision making has great advantages, the classic HRF should be extended to combine soft and hard failure of systems with condition monitoring. In this paper, we define the HRFs of hard and soft failure for system under condition monitoring and propose a method to estimate the HRFs with data of failure time. We discuss in detail the relationship between these two HRFs and the classical HRF. With double stochastic processes (processes of degradation and measured healthy status), the properties of these HRFs are also researched. Further the optimal maintenance decisions are made for non-repairable and repairable systems upon these two types of HRFs. Eventually, the idea of this paper is verified by numerical examples.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"109 1","pages":"962-968"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79125215","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8909025
Han Zhang, Jinglin Zhou, Jing Wang
Due to the existence of non-Gaussian interference in practical industry, more and more people are studying the parameter identification of non-Gaussian systems. Some results have been achieved. In this paper, a new double error entropy minimization algorithm is proposed for parameter identification of non-Gaussian systems. The algorithm is improved on the basis of the EDA algorithm and the Fixed-Point MEE algorithm. The simulation results show that the proposed algorithm can not only improve the accuracy of parameter identification but also estimate the required interference distribution. The algorithm has been successfully applied to the performance assessment of non-Gaussian systems and achieved good results.
{"title":"Performance Assessment of Non-Gaussian Systems Based on Double Error Entropy Minimization","authors":"Han Zhang, Jinglin Zhou, Jing Wang","doi":"10.1109/DDCLS.2019.8909025","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909025","url":null,"abstract":"Due to the existence of non-Gaussian interference in practical industry, more and more people are studying the parameter identification of non-Gaussian systems. Some results have been achieved. In this paper, a new double error entropy minimization algorithm is proposed for parameter identification of non-Gaussian systems. The algorithm is improved on the basis of the EDA algorithm and the Fixed-Point MEE algorithm. The simulation results show that the proposed algorithm can not only improve the accuracy of parameter identification but also estimate the required interference distribution. The algorithm has been successfully applied to the performance assessment of non-Gaussian systems and achieved good results.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"33 1","pages":"1177-1182"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88245038","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8908866
Jingkun Yan, Long Jin, Rui Zhang, Hongxin Li, Jiliang Zhang, Huiyan Lu
It is well known that many engineering communities are confronted with the problem of Lyapunov equation (LE) solving, and a mass of methods are proposed to solve the problem under the circumstance of no-noise. Whereas, noise is inevitable during actual experiments due to all kinds of factors, and models with resistance to noises are needed. To fill this lacuna, this paper proposes a zeroing-type recurrent neural network (ZTRNN) model for the task of solving the time-dependent LE. Additionally, the original zeroing neural network (OZNN) model is described for the purpose of comparison. Rigorous theoretical analyses concerning the convergence and the resistance to noises of the ZTRNN model are presented. Further, a numerical example of the time-dependent LE is solved by using the ZTRNN model and the OZNN model separately. Computer simulations are carried out perfectly and the results verify the feasibility and resistance to noises of the ZTRNN model, reflecting that the ZTRNN model outperforms the OZNN model in terms of resistance to noises.
{"title":"Zeroing-Type Recurrent Neural Network for Solving Time-Dependent Lyapunov Equation with Noise Rejection","authors":"Jingkun Yan, Long Jin, Rui Zhang, Hongxin Li, Jiliang Zhang, Huiyan Lu","doi":"10.1109/DDCLS.2019.8908866","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908866","url":null,"abstract":"It is well known that many engineering communities are confronted with the problem of Lyapunov equation (LE) solving, and a mass of methods are proposed to solve the problem under the circumstance of no-noise. Whereas, noise is inevitable during actual experiments due to all kinds of factors, and models with resistance to noises are needed. To fill this lacuna, this paper proposes a zeroing-type recurrent neural network (ZTRNN) model for the task of solving the time-dependent LE. Additionally, the original zeroing neural network (OZNN) model is described for the purpose of comparison. Rigorous theoretical analyses concerning the convergence and the resistance to noises of the ZTRNN model are presented. Further, a numerical example of the time-dependent LE is solved by using the ZTRNN model and the OZNN model separately. Computer simulations are carried out perfectly and the results verify the feasibility and resistance to noises of the ZTRNN model, reflecting that the ZTRNN model outperforms the OZNN model in terms of resistance to noises.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"44 1","pages":"366-371"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85627084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8909000
Shenglun Yi, X. Ren, Tingli Su
This paper considers an improved Kalman filter (KF) for a non-Gaussian system, when an adaptive statistics model is applied to capture the systematic characteristics in real time. The problem is formulated as self-adaptive adjustment parameters (SAPs) updating by the recursive least squares (RLS) algorithm. These parameters are shown to admit adaptive statistics model to characteristics of which applies and extends results given earlier in “Online denoising based on the second-order adaptive statistics model” (S. L. Yi and X. B. Jin et al., Sensors, 17(7), 1668, 2017.). Simulations comparing the improved KF based on the SAPs to the standard KF and the past algorithm illustrate a satisfactory performance when applied to self-adaptive adjustment parameters. Simulation results show that the proposed algorithm can gradually converge with a small performance loss.
本文研究了一种用于非高斯系统的改进卡尔曼滤波器(KF)，该滤波器采用自适应统计模型来实时捕捉系统特征。将该问题表述为用递归最小二乘(RLS)算法更新自适应调整参数。这些参数表明，自适应统计模型的特征适用并扩展了前面“基于二阶自适应统计模型的在线去噪”中给出的结果(S. L. Yi和X. B. Jin等人，传感器，17(7)，1668,2017)。将基于SAPs的改进KF算法与标准KF算法和过去的算法进行了仿真比较，结果表明，在自适应调整参数时，改进的KF算法具有令人满意的性能。仿真结果表明，该算法能以较小的性能损失逐步收敛。
{"title":"An Improved Kalman Filter Based on Self-adaptive Adjustment Parameters","authors":"Shenglun Yi, X. Ren, Tingli Su","doi":"10.1109/DDCLS.2019.8909000","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909000","url":null,"abstract":"This paper considers an improved Kalman filter (KF) for a non-Gaussian system, when an adaptive statistics model is applied to capture the systematic characteristics in real time. The problem is formulated as self-adaptive adjustment parameters (SAPs) updating by the recursive least squares (RLS) algorithm. These parameters are shown to admit adaptive statistics model to characteristics of which applies and extends results given earlier in “Online denoising based on the second-order adaptive statistics model” (S. L. Yi and X. B. Jin et al., Sensors, 17(7), 1668, 2017.). Simulations comparing the improved KF based on the SAPs to the standard KF and the past algorithm illustrate a satisfactory performance when applied to self-adaptive adjustment parameters. Simulation results show that the proposed algorithm can gradually converge with a small performance loss.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"8 1","pages":"1060-1064"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80725405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：