표제지
목차
국문요약 15
Ⅰ. 서론 16
1.1. 개요 16
1.2. 연구 동향 18
1.3. 연구의 내용 및 목적 21
Ⅱ. 로봇 시스템의 동적 모델링 및 특성분석 22
2.1. 모바일 로봇의 동적 모델링 및 특성분석 22
2.1.1. 모바일 로봇의 모델링 22
2.1.2. 동적 특성분석 및 안정성 분석 29
2.1.3. 모바일 로봇의 제어 구조 40
2.2. 다관절 로봇 매니퓰레이터의 동적 모델링 및 특성분석 46
2.2.1. 다관절 로봇 매니퓰레이터의 운동학 해석 46
2.2.2. 매니퓰레이터의 운동방정식 유도 56
2.2.3. 매니퓰레이터의 자코비안 해석 67
2.2.4. 구동 액추에이터 특성 분석 71
Ⅲ. 로봇의 작업지능제어 알고리즘 개발 75
3.1. 지능 제어기의 학습구조 77
3.1.1. 기본 학습구조 77
3.1.2. 지능 제어기법의 활성화 함수 특성분석 80
3.2. 로봇의 작업지능제어를 위한 반복학습 지능 제어기 구성 83
3.2.1. 기본이론 83
3.2.2. 반복학습제어 알고리즘 96
Ⅳ. 로봇 지능제어 실현을 위한 음성인식제어 102
4.1. 기본개요 102
4.2. 음성신호 처리 103
4.2.1. 잡음처리 103
4.2.2. 음성데이터 획득 105
4.2.3. 음성데이터의 특징추출 105
4.2.4. 벡터 양자화 108
4.3. 음성인식 알고리즘 112
4.3.1. 음성인식 알고리즘 112
4.4. 음성인식 제어시스템 116
4.4.1. 음성인식 시스템의 구성 116
4.4.2. 음성인식 제어 프로그램 개발 121
4.4.3. 음성인식 로봇 제어시스템 구조 124
Ⅴ. 성능실험 및 결과 129
5.1. 모바일 로봇의 성능 실험 129
5.1.1. 모바일 로봇의 구조 및 사양 129
5.1.2. 실험방법 및 조건 131
5.1.3. 모바일 로봇의 실험 및 결과 135
5.2. 다관절 로봇의 성능실험 및 결과 140
5.2.1. 6축 수직 다관절 로봇 시뮬레이션 성능실험 140
5.2.2. 제안된 제어기법의 학습제어 특성 및 결과 151
5.3. 음성명령 제어실험 및 결과 172
5.3.1. 모바일 로봇 음성명령제어 실험 및 결과 172
5.3.2. 다관절 로봇 매니퓰레이터 음성명령 제어실험 184
5.3.3. 구현된 음성 인식기를 이용한 보드와 로봇의 통신구현 187
5.3.4. 음성명령제어 성능실험 194
Ⅵ. 결론 199
참고문헌 201
ABSTRACT 209
Table. 2-1. Link parameter of D-H coordinate of 6-axis robot 48
Table. 4-1. Configuration of F1 layer 113
Table. 5-1. Parameter Data of the robot manipulator 142
Table. 5-2. Analysis of RMS error in cartesian coordinate system for trajectory 1 155
Table. 5-3. Analysis of RMS error in cartesian coordinate system for trajectory 2 160
Table. 5-4. Analysis of RMS error in cartesian coordinate system for trajectory 3 165
Table. 5-5. Voice recognition saved I/O input 191
Table. 5-6. Voice command recognition judgment result 198
Fig. 2-1. The coordinates system of mobile robot 22
Fig. 2-2. The travelling coordinates system of mobile robot 24
Fig. 2-3. Linear and angular velocity of mobile robot 41
Fig. 2-4. Link coordinates and structure of robot manipulator 46
Fig. 2-5. The coordinate of robot with six joints based D-H rule 48
Fig. 2-6. Diagram of robot about θ₁ 51
Fig. 2-7. Diagram of robot about θ₂~θ₃ 52
Fig. 2-8. Angular velocity and linear velocity 68
Fig. 2-9. Flowchart of the main program 73
Fig. 2-10. Flowchart of the interrupt service routine 74
Fig. 2-11. Flowchart of control input 74
Fig. 3-1. Multi-layer neural network 77
Fig. 3-2. Activation function 81
Fig. 3-3. The parameter estimation 84
Fig. 3-4. The scheme of off-line neural controller 85
Fig. 3-5. The scheme of on-line neural controller 87
Fig. 3-6. McCulloch-Pitts neuron model 89
Fig. 3-7. Delta learning rule 91
Fig. 3-8. The diagram of neural network control system 92
Fig. 3-9. The structure of multi-layer perceptron neural network based on pattern recognition 98
Fig. 3-10. The structure of pattern recognition perceptron 99
Fig. 3-11. The structure of proposed neural network control system 101
Fig. 4-1. Spectrum subtraction method 104
Fig. 4-2. Produce of MFCC process 105
Fig. 4-3. Bandwidth of Mel-Cepstrum(Bandwidth of triangle filter) 108
Fig. 4-4. Processing of vector quantization 109
Fig. 4-5. Procedure of binary tree algorithm 111
Fig. 4-6. Speech recognition algorithm structure 112
Fig. 4-7. Speech recognition algorithm flowchart 115
Fig. 4-8. Configuration of speech recognition board 116
Fig. 4-9. Speech data input/output part 117
Fig. 4-10. Voice signal processing unit using digital signal processor 118
Fig. 4-11. Memory map 119
Fig. 4-12. I/O map 120
Fig. 4-13. System interface part 120
Fig. 4-14. Speech recognition system structure 121
Fig. 4-15. Region detection of real speech 122
Fig. 4-16. Voice recognition method 123
Fig. 4-17. Error range 124
Fig. 4-18. Control system configuration of voice recognition robot 125
Fig. 4-19. Power part 126
Fig. 4-20. Configuration of main system 127
Fig. 4-21. Ultra sensor circuit 127
Fig. 4-22. Total system 128
Fig. 5-1. The structure and specification of mobile robot 129
Fig. 5-2. Straight path and learning outcomes 132
Fig. 5-3. Change in x value on a straight path 132
Fig. 5-4. X-axis error on a straight path 133
Fig. 5-5. Change of y value on straight path 133
Fig. 5-6. Y-axis error on a straight path 134
Fig. 5-7. Change of θ value on a straight path 134
Fig. 5-8. θ error on straight path 135
Fig. 5-9. The real model of mobile robot 136
Fig. 5-10. Practical experiment: results on a straight path 137
Fig. 5-11. Practical experiment: variation of x-values on a straight path 138
Fig. 5-12. Practical experiment: x-axis errors on a straight path 138
Fig. 5-13. Practical experiment: variation of y-values on a straight path 139
Fig. 5-14. Practical experiment: y-axis error on a straight path 139
Fig. 5-15. 6-axis vertical articulated robot structure 140
Fig. 5-16. The initial screen of the 3D simulation program 141
Fig. 5-17. Generation of 6 link parameters 141
Fig. 5-18. 3D simulation operation 144
Fig. 5-19. Joint 1 performance analysis and data verification 145
Fig. 5-20. Joint 2 performance analysis and data verification 146
Fig. 5-21. Joint 3 performance analysis and data verification 147
Fig. 5-22. Joint 4 performance analysis and data verification 148
Fig. 5-23. Joint 5 performance analysis and data verification 149
Fig. 5-24. Joint 6 performance analysis and data verification 150
Fig. 5-25. 6-axis vertical articulated robot 151
Fig. 5-26. Result of trajectory tracking control of trajectory 1 by the proposed neural controller 154
Fig. 5-27. Result of trajectory tracking control of trajectory 1 by PD controller 155
Fig. 5-28. Result of trajectory tracking of velocity trajectory 2 by the proposed neural controller 157
Fig. 5-29. The error of trajectory 2 by the proposed neural controller 158
Fig. 5-30. Result of trajectory tracking of velocity trajectory 2 by PD controller 159
Fig. 5-31. The error of trajectory 2 by PD controller 160
Fig. 5-32. Trajectory tracking result of trajectory 3 by the proposed neural controller 161
Fig. 5-33. The error of trajectory 3 by the proposed neural controller 162
Fig. 5-34. Trajectory tracking result of trajectory 3 by PD controller 163
Fig. 5-35. The error of trajectory 3 by PD controller 164
Fig. 5-36. Speed, torque, and position test results of joint 1 of the 6-axis articulated robot 166
Fig. 5-37. Speed, torque, and position test results of joint 2 of the 6-axis articulated robot 167
Fig. 5-38. Speed, torque, and position test results of joint 3 of the 6-axis articulated robot 168
Fig. 5-39. Speed, torque, and position test results of joint 4 of the 6-axis articulated robot 169
Fig. 5-40. Speed, torque, and position test results of joint 5 of the 6-axis articulated robot 170
Fig. 5-41. Speed, torque, and position test results of joint 6 of the 6-axis articulated robot 171
Fig. 5-42. Voice command recognition board structure 173
Fig. 5-43. Voice command recognition program setting 176
Fig. 5-44. Voice command recognition program execution screen 177
Fig. 5-45. Scenario action editor 179
Fig. 5-46. Scenario action editor program test 179
Fig. 5-47. Mobile robot voice command recognition waveform results 180
Fig. 5-48. Voice command recognition performance test I 181
Fig. 5-49. Continuous sentence speech command recognition test result 182
Fig. 5-50. Basic operation motion voice command recognition test result 182
Fig. 5-51. Voice command recognition performance test II 183
Fig. 5-52. Voice command recognition program test 185
Fig. 5-53. Voice recognition module of SpeakUp Click 186
Fig. 5-54. UART output data of SpeakUp Click(Data from Index 0(00) to F(15)) 187
Fig. 5-55. Configuration of the voice command recognition robot control system 188
Fig. 5-56. Control module of voice command recognition 189
Fig. 5-57. Voice command recognition execution program 190
Fig. 5-58. I/O setting screen for recorded voice 190
Fig. 5-59. Learning and testing procedure of voice command recognition system 193
Fig. 5-60. Voice command recognition waveform result of articulated robot 194
Fig. 5-61. Voice command recognition structure 195
Fig. 5-62. The real model of voice command control experiment 196
Fig. 5-63. Voice command control test result 197