Pub Date : 2019-05-01DOI: 10.1109/DDCLS.2019.8909073
Qiao Zhu, Jun-Xiong Chen, Fazhi Song, Yang Liu
In this work, we propose a new and simple controller design based on internal model principle (IMP) and existing PID, with the implementation on a wafer stage. PID has been well applied to solve control engineering problems owing to its flexibility to handle various control tasks, and the easiness to tune PID gains for a variety of plants. However, PID may fail to achieve satisfactory tracking performance when dealing with a complex control task that is beyond a simple step reference. IMP suggests that a good control response can only be expected when the internal model of the reference is incorporated into the control loop. For a well-tuned PID, however, it is not that straightforward to incorporate the internal model of the reference into the existing PID. In this work, focusing on a real wafer stage, we explore a new control design approach, reference-model (RM)-based PID, to achieve the internal model and meanwhile retain the well-tuned PID. The new controller design for the wafer stage is achieved through several steps. First, the body plots of the experimental setup of wafer stage are acquired and a second-order approximate model is established. Second, a reference model specific in wafer stage is given and its inherent internal model is analyzed. Then, the feedback controller is constructed by concatenating the internal model and an existing PID that has been well tuned. Next, considering the low-frequency characteristic of the reference, the internal model is further modified with a simple high-pass filter such that more of high-frequency components in the tracking error could be suppressed. Finally, in experiments, more accurate tracking performance of the RM-based PID is illustrated by comparing with two classic PIDs with different bandwidths.
{"title":"A Simple Reference Model Based PID Control for Wafer Scanner Systems","authors":"Qiao Zhu, Jun-Xiong Chen, Fazhi Song, Yang Liu","doi":"10.1109/DDCLS.2019.8909073","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909073","url":null,"abstract":"In this work, we propose a new and simple controller design based on internal model principle (IMP) and existing PID, with the implementation on a wafer stage. PID has been well applied to solve control engineering problems owing to its flexibility to handle various control tasks, and the easiness to tune PID gains for a variety of plants. However, PID may fail to achieve satisfactory tracking performance when dealing with a complex control task that is beyond a simple step reference. IMP suggests that a good control response can only be expected when the internal model of the reference is incorporated into the control loop. For a well-tuned PID, however, it is not that straightforward to incorporate the internal model of the reference into the existing PID. In this work, focusing on a real wafer stage, we explore a new control design approach, reference-model (RM)-based PID, to achieve the internal model and meanwhile retain the well-tuned PID. The new controller design for the wafer stage is achieved through several steps. First, the body plots of the experimental setup of wafer stage are acquired and a second-order approximate model is established. Second, a reference model specific in wafer stage is given and its inherent internal model is analyzed. Then, the feedback controller is constructed by concatenating the internal model and an existing PID that has been well tuned. Next, considering the low-frequency characteristic of the reference, the internal model is further modified with a simple high-pass filter such that more of high-frequency components in the tracking error could be suppressed. Finally, in experiments, more accurate tracking performance of the RM-based PID is illustrated by comparing with two classic PIDs with different bandwidths.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"33 1","pages":"99-104"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87931786","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.2019.8908856
Yunyun Jin, Yang Song, Taicheng Yang, Weiyan Hou, M. Schüller
This paper considers the global exponential stabilizability (GES) of a switched linear system under language constraints, which can be described by a nondeterministic finite state automaton. Firstly, the automaton is represented as a labeled diagraph to reduce the problem to the GES analysis in strongly connected components. Secondly, we analysis the properties of the lifted labeled diagraph, which can express the dwell time constraints intuitively. Based on the lifted labeled diagraph, we generalize the Lyapunov-Metzler condition to an M-step version, and propose a less conservative condition based on S-procedure. Finally, a numerical example is provided to demonstrate the S-procedure condition.
{"title":"Exponential Stabilizability Analysis for Constrained Switched System","authors":"Yunyun Jin, Yang Song, Taicheng Yang, Weiyan Hou, M. Schüller","doi":"10.1109/DDCLS.2019.8908856","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908856","url":null,"abstract":"This paper considers the global exponential stabilizability (GES) of a switched linear system under language constraints, which can be described by a nondeterministic finite state automaton. Firstly, the automaton is represented as a labeled diagraph to reduce the problem to the GES analysis in strongly connected components. Secondly, we analysis the properties of the lifted labeled diagraph, which can express the dwell time constraints intuitively. Based on the lifted labeled diagraph, we generalize the Lyapunov-Metzler condition to an M-step version, and propose a less conservative condition based on S-procedure. Finally, a numerical example is provided to demonstrate the S-procedure condition.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"88 1","pages":"934-938"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89050100","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.2019.8908943
Pengfei Xia, Wei Wei, Zaiwen Liu, Min Zuo
Positioning control of a nano-positioner driven by a piezoelectric actuator is discussed. Robust adaptive control with extended state observer is presented for the trajectory tracking control. Radial basis function neural network (RBFNN) is utilized to estimate the unknown nonlinearities. Extended state observer (ESO) is also taken to observe the total disturbance, which includes external disturbances and hysteresis. Both the RBFNN and the ESO are utilized to reduce the dependence on model information. A nano-positioner model is established. Simulations confirm the robust adaptive control with ESO is effective in improving positioning accuracy.
{"title":"A Robust Adaptive Control with Extended State Observer for a Piezo-actuated Nano-positioner","authors":"Pengfei Xia, Wei Wei, Zaiwen Liu, Min Zuo","doi":"10.1109/DDCLS.2019.8908943","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908943","url":null,"abstract":"Positioning control of a nano-positioner driven by a piezoelectric actuator is discussed. Robust adaptive control with extended state observer is presented for the trajectory tracking control. Radial basis function neural network (RBFNN) is utilized to estimate the unknown nonlinearities. Extended state observer (ESO) is also taken to observe the total disturbance, which includes external disturbances and hysteresis. Both the RBFNN and the ESO are utilized to reduce the dependence on model information. A nano-positioner model is established. Simulations confirm the robust adaptive control with ESO is effective in improving positioning accuracy.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"64 1","pages":"1003-1007"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83775791","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.2019.8909032
Saim Ahmed, Haoping Wang, Yang Tian
Generally model-reference adaptive control (MRAC) is designed using known regression matrix. However, the formulation of regression matrix is difficult for more degree of freedoms (DOFs) of robot manipulator and sometime impossible to compute for many applications. In this work, MRAC using time delay estimation (TDE) named (MRAC-TDE) is proposed to avoid complex calculation of regression matrix and provides model-free control. Therefore, TDE is devised to estimate the unknown dynamics and MRAC is used to update the control gains. The closed-loop stability of system is investigated using the Lyapunov stability criterion. In the end, to validate the effectiveness of the proposed method, simulations are illustrated the appropriateness of proposed MRAC-TDE.
{"title":"Time Delay Estimation Based Model Reference Adaptive Control for Robot Manipulators","authors":"Saim Ahmed, Haoping Wang, Yang Tian","doi":"10.1109/DDCLS.2019.8909032","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909032","url":null,"abstract":"Generally model-reference adaptive control (MRAC) is designed using known regression matrix. However, the formulation of regression matrix is difficult for more degree of freedoms (DOFs) of robot manipulator and sometime impossible to compute for many applications. In this work, MRAC using time delay estimation (TDE) named (MRAC-TDE) is proposed to avoid complex calculation of regression matrix and provides model-free control. Therefore, TDE is devised to estimate the unknown dynamics and MRAC is used to update the control gains. The closed-loop stability of system is investigated using the Lyapunov stability criterion. In the end, to validate the effectiveness of the proposed method, simulations are illustrated the appropriateness of proposed MRAC-TDE.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"95 1","pages":"261-265"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83919698","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.2019.8909058
Lingyu Zhang, Dehui Sun, Li Wang, Haibo Zhang
Aiming at the problems of heavy workload of basic data collection, complicated manual parameters calibration and inaccurate calibration in the traditional microscopic traffic simulation model parameter calibration, an adaptive microscopic traffic simulation model parameter calibration method based on floating car data is proposed. First, the basic data of the simulation road network were obtained by using the floating car technology. Secondly, the parameters calibration process of the microscopic traffic simulation model was constructed, and the self- adaptive orthogonal genetic was used to achieve the model parameters calibration. Finally, using the actual data of the South Ring Road main line, District Changping, Beijing to simulate and verify. The results show that the proposed method in this paper can not only reduce the workload of manual calibration, but also the model parameter calibration is more accurate, which proves the feasibility and effectiveness of the method.
{"title":"Parameter Calibration of Microscopic Traffic Simulation Model Based on Floating Car Data","authors":"Lingyu Zhang, Dehui Sun, Li Wang, Haibo Zhang","doi":"10.1109/DDCLS.2019.8909058","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909058","url":null,"abstract":"Aiming at the problems of heavy workload of basic data collection, complicated manual parameters calibration and inaccurate calibration in the traditional microscopic traffic simulation model parameter calibration, an adaptive microscopic traffic simulation model parameter calibration method based on floating car data is proposed. First, the basic data of the simulation road network were obtained by using the floating car technology. Secondly, the parameters calibration process of the microscopic traffic simulation model was constructed, and the self- adaptive orthogonal genetic was used to achieve the model parameters calibration. Finally, using the actual data of the South Ring Road main line, District Changping, Beijing to simulate and verify. The results show that the proposed method in this paper can not only reduce the workload of manual calibration, but also the model parameter calibration is more accurate, which proves the feasibility and effectiveness of the method.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"255 1","pages":"1219-1224"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78216164","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}
With the advent of online social networks, the approach of recommendation based on social network has emerged. However, some recommendation algorithms based on the trust network do not fully mine the information of user's trust relationships. To alleviate such problems, we propose a socialRT method, which is a social recommendation trust method based on joint matrix decomposition. The proposed socialRT method collective factorizes the following relationship matrix and the social trust relationship matrix to obtain the recommendation model. We have conducted experiments on Sina Weibo dataset, the experimental results demonstrate that the proposed recommendation method leads to a substantial increase in recommendation quality.
{"title":"SocialRT: A Recommendation Method Based On Social Trust","authors":"Xing Xing, Zhixin Meng, Hanting Chu, Jianyan Luo, Tiansheng Qu, Zhichun Jia","doi":"10.1109/DDCLS.2019.8908972","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908972","url":null,"abstract":"With the advent of online social networks, the approach of recommendation based on social network has emerged. However, some recommendation algorithms based on the trust network do not fully mine the information of user's trust relationships. To alleviate such problems, we propose a socialRT method, which is a social recommendation trust method based on joint matrix decomposition. The proposed socialRT method collective factorizes the following relationship matrix and the social trust relationship matrix to obtain the recommendation model. We have conducted experiments on Sina Weibo dataset, the experimental results demonstrate that the proposed recommendation method leads to a substantial increase in recommendation quality.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"12 1","pages":"443-447"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78354847","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.2019.8909080
Yuwei Zhang, Ke Deng, Yongcai Pan, Qingzheng Liu
In view of the inadequacy of Itti's computational model of visual attention mechanism in significance region detection, an improved computational model of visual attention based on edge detection was proposed here. It is based on the human eye's perception advantage of the edge shape information of the target object. On the basis of Itti model, this model can improve the extraction effect of significant regions in visual attention computing model by introducing edge information, and can segment significant regions more accurately. The experiment shows that the success rate of target object contour detection in this method can reach 91%, which is higher than the traditional detection method in calculation speed, and the target object contour recognition effect is better.
{"title":"A Visual Attention Computational Model Based on Edge Detection","authors":"Yuwei Zhang, Ke Deng, Yongcai Pan, Qingzheng Liu","doi":"10.1109/DDCLS.2019.8909080","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909080","url":null,"abstract":"In view of the inadequacy of Itti's computational model of visual attention mechanism in significance region detection, an improved computational model of visual attention based on edge detection was proposed here. It is based on the human eye's perception advantage of the edge shape information of the target object. On the basis of Itti model, this model can improve the extraction effect of significant regions in visual attention computing model by introducing edge information, and can segment significant regions more accurately. The experiment shows that the success rate of target object contour detection in this method can reach 91%, which is higher than the traditional detection method in calculation speed, and the target object contour recognition effect is better.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"16 1","pages":"957-961"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78363291","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.2019.8909074
Zeqing Ma, Jinrui Ren, Yi-Wei Lin, Q. Quan
In this paper, an auxiliary modeling method for system identification is proposed for multicopter dynamic modeling. By utilizing an auxiliary controller called damper during the modeling and identification process, the unstable attitude angle and angular rate channels of multicopters can turn to be stable so as to obtain the parameterized dynamic model without safety problems led by traditional methods and large-space requirement. Through an application to an indoor fixed quadcopter system, simulation results demonstrate the feasibility of the proposed method for the multicopter dynamic modeling.
{"title":"An Auxiliary Model Construction Method for System Identification and Its Application to An Indoor Multicopter Platform","authors":"Zeqing Ma, Jinrui Ren, Yi-Wei Lin, Q. Quan","doi":"10.1109/DDCLS.2019.8909074","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8909074","url":null,"abstract":"In this paper, an auxiliary modeling method for system identification is proposed for multicopter dynamic modeling. By utilizing an auxiliary controller called damper during the modeling and identification process, the unstable attitude angle and angular rate channels of multicopters can turn to be stable so as to obtain the parameterized dynamic model without safety problems led by traditional methods and large-space requirement. Through an application to an indoor fixed quadcopter system, simulation results demonstrate the feasibility of the proposed method for the multicopter dynamic modeling.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"2 1","pages":"1189-1195"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78449116","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.2019.8908958
Yi Liu, Jiu-sun Zeng, Lei Xie, Xun Lang, Shihua Luo, H. Su
This paper focus on developing an effective method to monitor the industrial process with multiple operation conditions. By utilizing the technique of probabilistic linear discriminant analysis (PLDA), the between- and within-class latent variables can extract more useful information. The proposed method, the modified PLDA (MPLDA), transforms the centralized samples into a new type of between-class latent variables. The current mode operation condition can be identified by comparing a series of cosine similarities deduced by the original and the new between-class latent variables. The online monitoring procedures are built on the basis of this mode identification. Unlike the conventional $T^{2}$ and $Q$ statistics designed for within-class latent variable, the proposed monitoring statistics take both between- and within-class latent variables into consideration. For the model training, the joint updating expectation-maximization (EM) algorithm is developed. The enhanced performance of the MPLDA based method is illustrated by the application of Tennessee Eastman (TE) process.
{"title":"Multimode Process Monitoring Based on Modified Probabilistic Linear Discriminant Analysis","authors":"Yi Liu, Jiu-sun Zeng, Lei Xie, Xun Lang, Shihua Luo, H. Su","doi":"10.1109/DDCLS.2019.8908958","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908958","url":null,"abstract":"This paper focus on developing an effective method to monitor the industrial process with multiple operation conditions. By utilizing the technique of probabilistic linear discriminant analysis (PLDA), the between- and within-class latent variables can extract more useful information. The proposed method, the modified PLDA (MPLDA), transforms the centralized samples into a new type of between-class latent variables. The current mode operation condition can be identified by comparing a series of cosine similarities deduced by the original and the new between-class latent variables. The online monitoring procedures are built on the basis of this mode identification. Unlike the conventional $T^{2}$ and $Q$ statistics designed for within-class latent variable, the proposed monitoring statistics take both between- and within-class latent variables into consideration. For the model training, the joint updating expectation-maximization (EM) algorithm is developed. The enhanced performance of the MPLDA based method is illustrated by the application of Tennessee Eastman (TE) process.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"49 1","pages":"604-609"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85442785","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.2019.8908850
Feifan Shen, Lingjian Ye, Saite Fan, Zhiqiang Ge, Zhihuan Song
This paper handles with the problem of the run-to-run trajectory prediction of batch processes with uneven batch length. Most current data-driven works focus on the run-to-run variations during both batch trajectory modeling and prediction stages. However, batch-to-batch correlations should be drawn extreme attentions to when gradual changes exist in batch sequence. To obtain a better batch trajectory prediction performance of uneven-length batch processes, dynamic time warping (DTW) and long-short term memory (LSTM) neural network are introduced in this work to extract batch-to-batch correlations. Firstly, the recursive DTW is used to synchronize uneven batch samples. Then, the LSTM neural network is introduced to extract the run-to-run batch correlations during the trajectory modeling stage. Finally, online batch trajectory prediction can be implemented according to the offline LSTM model. A simulation based on the fed-batch penicillin fermentation process is provided to testify the effectiveness of the proposed method.
{"title":"Run-to-run Trajectory Prediction of Uneven-length Batch Processes Using DTW-LSTM","authors":"Feifan Shen, Lingjian Ye, Saite Fan, Zhiqiang Ge, Zhihuan Song","doi":"10.1109/DDCLS.2019.8908850","DOIUrl":"https://doi.org/10.1109/DDCLS.2019.8908850","url":null,"abstract":"This paper handles with the problem of the run-to-run trajectory prediction of batch processes with uneven batch length. Most current data-driven works focus on the run-to-run variations during both batch trajectory modeling and prediction stages. However, batch-to-batch correlations should be drawn extreme attentions to when gradual changes exist in batch sequence. To obtain a better batch trajectory prediction performance of uneven-length batch processes, dynamic time warping (DTW) and long-short term memory (LSTM) neural network are introduced in this work to extract batch-to-batch correlations. Firstly, the recursive DTW is used to synchronize uneven batch samples. Then, the LSTM neural network is introduced to extract the run-to-run batch correlations during the trajectory modeling stage. Finally, online batch trajectory prediction can be implemented according to the offline LSTM model. A simulation based on the fed-batch penicillin fermentation process is provided to testify the effectiveness of the proposed method.","PeriodicalId":6699,"journal":{"name":"2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS)","volume":"17 1","pages":"1183-1188"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86929132","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：