크로스벨트 분류기의 인덕션 컨베이어라인용 속도 제어기의 개발
- 발행기관 부경대학교
- 지도교수 Kim Sang Bong
- 발행년도 2018
- 학위수여년월 2018.2
- 학위명 박사
- 학과 및 전공 대학원 기계설계공학과
- 실제URI http://www.dcollection.net/handler/pknu/200000010858
- UCI I804:21031-200000010858
- 본문언어 영어
- 저작권 부경대학교 논문은 저작권에 의해 보호받습니다.
초록/요약
Nowadays, cross-belt sorter systems are widely utilized in distribution centers. Since their speed are fast with approximately 2m/s and more, loading accurately parcels on vacant carriers is a challenge. Controlling conveyor velocities to track desired velocity profiles accurately is necessary in order to do this task. This dissertation presents development results of belt conveyor velocity controllers for an induction conveyor line in a cross-belt sorter system. To shoot a parcel into a vacant carrier of the loop cross-belt conveyor system, the main control subsystem utilized by a industrial computer ARK 5260 detects the parcel on the ICL by a scanner and the suitable vacant carrier by UV sensors attached in the loop cross-belt conveyor subsystem. The main control subsystem creates velocity profiles for conveyors of the ICL and sends them to the velocity controller implemented by TMS320F28335 DSP microprocessor of the ICL. The velocity profiles play a role as the velocity reference inputs. The velocity controller of the ICL should control conveyor velocities measured by encoders to track these velocity reference inputs. To do this task, the followings are done. Firstly, the cross-belt sorter system description and system modeling of an induction conveyor line subsystem consisting of an on-loading conveyor, a buffer conveyor and a transition conveyor are presented. Each conveyor is developed with mechanical and electrical parts such as driving roller, driven roller, encoder, induction motor, etc. The system modeling of the induction conveyor line subsystem without/with belt elasticity describes the behaviors of the mechanical and electrical operating mechanism to obtain a dynamic system expressed by state equations. Secondly, a velocity controller with input saturation for the ICL using a conventional model reference adaptive control method is designed. A reference model with the time derivative of reference inputs for an output vector of the reference model to track asymptotically a trapezoidal reference input vector is proposed. A control input signal saturation can lead to a poor control performance and even a closed loop instability unless its effect is considered in a design of the controller. Hence, an auxiliary error vector is added in a state error dynamics to compensate for the effect of the input control signal saturation. The auxiliary error vector guarantees that it goes to zero when the input control signal saturation does not take place and it is bounded when the input control signal saturation takes place. This means that the closed-loop control system is locally stable. Experimental results are shown to verify effectiveness of the proposed conventional model reference adaptive controller (MRAC) for input saturation. Thirdly, although the MRAC with a sufficient adaptive controller gain can achieve asymptotic tracking, the time for the designed MRAC achieving the asymptotic tracking is extremely long due to a large deviation of the state error in transient state. Hence, fast adaptation is needed in order to improve tracking performance by utilizing a large adaptive gain to reduce the state error rapidly. However, if the adaptive gain increases, unexpected high-frequency elements are generated in a control input signal. To eliminate the unexpected high-frequency elements in the control input signal, a reference model modified by adding a term being product of the state error and a feedback error gain is proposed. A new velocity controller utilizing the modified reference model is proposed and named as modified-model reference adaptive controller (M--MRAC). In addition, e-modification is also utilized in update laws to be robust for the proposed controller under the presence of bounded disturbances. Simulation and Experimental results are shown to verify effectiveness of the proposed controller for eliminating unexpected high-frequency elements by comparing with that of the MRAC. Fourthly, the "parameters drift" phenomenon caused by bounded disturbance was eliminated by update laws with e-modification. However, for large state error, e-modification slows down adaptation process and this undesirable effect contradicts the control object of reducing the state error as fast as possible. A projection operator is addressed to fast adaptation uncertain parameters and at the same time the projection operator enforces uniform boundedness of estimated parameters. A new M--MRAC utilized the projection operator in update laws is proposed to obtain fast adaptation and being robust the proposed controller under the presence of bounded disturbance. In addition, to eliminate the unexpected high-frequency elements in the control input signal, the feedback error gain has to increase. As a result, angular velocity outputs cannot track reference inputs since a reference model error increases if the feedback error gain is too large. Hence, the constraint of state error, feedback error gain and adaptive gain are presented. Simulation and experimental results are shown to verify effectiveness of the proposed controller for fast adaptation by projection operator and eliminating unexpected high-frequency elements with the suitable feedback error gain.
more목차
Chapter 1. Introduction 1
1.1 Background and motivation 1
1.2 Problem Statements 7
1.3 Objective and researching method 8
1.4 Outline of dissertation and summary of contributions 10
Chapter 2. Induction Conveyor Line Description and
Modeling 13
2.1 System description 14
2.1.1 Cross-belt sorter system 14
2.1.2 ICL description 15
2.1.2.1 Mechanical design 16
2.1.2.2 Electrical design 19
2.2 ICL modeling 23
2.2.1 Modeling without belt elasticity 24
2.2.2 Modeling with belt elasticity 27
Chapter 3. Velocity controller design using MRAC with
input saturation 30
3.1 Model reference adaptive controller with input
saturation 30
3.2 Experimental results 39
3.3 Summary 53
Chapter 4. Velocity controller design using M-MRAC
for tracking improvement in transient phase... 55
4.1 Modified model reference adaptive controller (M–
MRAC) design 55
4.2 Simulation results 63
4.3 Experimental results 75
4.4 Summary 82
Chapter 5. Velocity controller design using M-MRAC
and projection operator in presence of
bounded disturbances 86
5.1 Projection operator and its properties 86
5.2 M-MRAC controller design using projection operator 91
5.2.1 Modified model reference adaptive controller
(M-MRAC) 92
5.2.2 Constraint of the modeling error vector, the
feedback gain and the adaptation rate 97
5.3 Simulation results 99
5.4 Summanry 115
Chapter 6. Conclusions and Future Works 118
6.1 Conclusions 118
6.2 Future works 122
References 123
Publication and Conference 133
Chapter A. Appendix A. The proof of Eqs. (3.21) and
(3.25) 140
Chapter B. Appendix B. The proof of Eq. (4.14)) and (4.18) 143
Chapter C. Appendix C. The proof of Eq. (5.29) 145