摘要
随着机器人向仿人型、智能化、高度集成的方向发展,微型六维力/力矩传感器作为提高机器人灵巧操作水平和智能化程度的重要部件之一,显得越来越重要。因此,本文在两项国家“863”项目(“基于MEMS的微型六维力/力矩传感器的研究”课题编号:2002AA423220;和“基于薄膜技术的全平面微型传感器的研究”课题编号:2003AA404220)的资助下,重点研究了微型六维力/力矩传感器及其自动标定技术。
DLR/HIT仿人灵巧手由四个相同结构的手指组成,采用机电一体化设计思想,将机械本体与传感、驱动、传动以及电气系统等一体化设计,实现系统的高度集成。它要求各类传感器信号本地数字化,并具有微小型尺寸的特点。因此,作为灵巧手指尖应用的六维力/力矩传感器,对其微型化、数字化集成及自动标定功能提出了较高的要求。本文正是围绕上述几点要求开展研究的,并主要分为以下几方面内容:传感器的研制、传感器微型化关键技术及MEMS薄膜工艺的研究,多维力传感器静动态性能研究、多维力传感器自动标定的研究。
首先,本文进行了六维力/力矩传感器弹性体的设计,针对传感器微型化的需要,阐述了两种微型化的传感器弹性体(薄壁圆筒形弹性体和全平面型弹性体)的结构形式和应变测量原理。针对薄壁圆筒形弹性体本文提出了一种集成式的电阻应变计,阐述了它与弹性体组成微型六维力/力矩传感器敏感部件的方法,及其设计和制作过程。同时,本文设计了六维力传感器的微型信号调理放大电路和基于数字信号处理器的微型数字化电路。并最终完成了具有微型化尺寸的(外径Φ16.5mm,高度H17.5mm)内嵌数字信号处理器的六维力/力矩传感器,从作者文献检索来看,尚未见到尺寸达到此等水平的六维力/力矩传感器。
针对全平面弹性体的结构特点,本文提出了采用薄膜工艺在铝合金为基底材料上制作薄膜电阻应变计,设计并制作了微型六维力/力矩传感器所用的薄膜电阻应变计结构。在各功能材料的分析与选择中,本文针对铝合金材料的特点,着重对四种绝缘层材料进行了实验,实验结果表明选择Al2O3做为绝缘层材料具有较好的效果。采用了引线键合技术将应变计组成的测量电桥信号引到信号调理电路板,提高了引线的可靠性,减少了引线工作量。本文对所制作的薄膜电阻应变计各层厚度进行了测量,并提出了采用原子力显微镜对薄膜厚度进行直接测量的方法。实际效果表明上述各种工艺的实施为六维力/力矩传感器的微型化的实现,系统可靠性的提高提供了有力的保障。并最终研制出基于薄膜技术的微型六维力/力矩传感器,实现了机电集成和数字化。
本文对所研制的微型六维力/力矩传感器进行了静动态性能的研究。通过静态加载实验得到了六维力/力矩传感器的静态输出特性及耦合情况。并深入研究了六维力传感器的静态解耦方法,提出了采用神经网络进行静态解耦的方法,以解决六维力传感器的表现出一定的非线性特性。通过对标定矩阵的扰动分析,确定了标定矩阵C的扰动引起的测量力F的误差上限,并最终给出传感器的各项静态性能指标。通过对传感器动态实验建模给出传感器动态性能指标,并根据传感器动态特性采用极点配置法设计了动态补偿器,并给出动态补偿仿真结果。
本文从两个方面研究了六维力/力矩传感器的自动标定问题:一是简化传统标定多维力传感器多次加载的过程。本文提出了基于三轴加速度计的多维力/力矩传感器自动标定的方法。将三轴加速度计静态应用,可以获得重力加速度相对于传感器坐标的三个倾角。根据此三个倾角计算砝码力,并自动获得砝码作用在传感器上的载荷。此过程无需多次施加砝码,并根据传感器的输出自动求解传感器的标定矩阵。二是采用“从运动复原形状”方法,分别实现了从二维力、三维力、二维力一维力矩及六维力/力矩传感器的自动标定。进而实现了多维力传感器的联合自动标定。该方法通过在多维力传感上任意施加力,并记录传感器的输出,便可恢复出施加在传感器上的力及标定矩阵的伪逆。不仅提高了标定的效率,而且,由于避免了传感器加载引起的输入误差,从而提高了标定精度。
The miniature six axis force/torque sensor is more and more important, as the robot development towards humanoid, high integrated, intelligentized. The miniature six axis force/torque sensor is one of important component of a robot, which can improve the overall level of the robot’s dexterous manipulation and intelligence level. This dissertation is supported by the the High Technology Research and Development Programme of China (“Study on the miniature six axis force/torque sensor based on the MEMS”No.2003AA404220 &“Study on the full surface sensor based on the thin film technology”No. 2002AA423220). The main research contents of this dissertation are the miniature six axis force/torque sensor and its auto-calibration technology.
The DLR/HIT dexterous hand is composed of four identical fingers with same structure, which is designed with idea of high integration system based on mechanical structure, sensing, driving, transmission and electronics. The integration system needs every sensor is digital locally, and small size. So, as the finger-tip sensor, the six force/torque sensor should be a small one and digital integration sensor. The ability of auto-calibration is plus. This dissetation focuses on these demands and carries on research works. The main contents are as follows: fabrication of the sensor, study on the key technology of miniaturization of the sensor and fabrication processes based on the MEMS thin film, study on the static & dynamic characteristics of the force sensor, and study on the auto-calibration of the multi-axis force sensor.
First, a process of six axis force/torque sensor’s sensing element design is presented. For the sensor’s micromation, the structure and measurement theory of two types of miniature sensing elements(the thin shell cylinder type of sensing element & the full surface type of sensing element) are presented. A kind of integrated strain gauge is presented, which is fit for the thin shell cylinder type of sensing element to compose the miniature six axis force sensor. A miniature signal conditioning circuit & DSP circuit are also presented. The sensor has a miniature size(16.5mm in diameter, 17.5mm in height) & a miniature embeded DSP circuit. According to the author’s references finding, no one six axis force/torque sensor has such small size existed.
According to the characteristics of the thin shell cylinder type of sensing element, the process of fabricating the thin film strain gauge on the aluminum sensing element is presented. The structure of the thin film strain gauge is designed and fabricated. By analysing and chosing every materials of function layers, this dissertation carries out experiments on four kinds of material of isolation layer, which fit for the aluminum sensing element. The results reveal that the Al2O3 is the best material, for its good performances as the isolation layer. The wire bonding is adopted to introduce the signals of the measurment Wheatstone bridges into the singnal conditioning circuit board, which improves the reliability of the circuit board and simplifies the wiring works. Every layer of the thin film strain gauge is measured.
A method of measuring the thickness of the thin film by the AFM is presented.The results proved that the processes above are feasible and powerful backup for the sensor system realization. Finally, the miniature six axis force/torque sensor based on the thin film technology is fabricated, which realized integration and digital output. This dissertation have conducted investigations in the static & dynamic characteristics of the miniature six axis force/torque sensor. The static outputs and coupling characteristics are outlined by appling static loads. A further study on the static decoupling methods have been done. Using the neural network as a new approch to improve the decoupling performance is presented, which can overcome some nonlinear characteristics of the sensor. By the calibration matrix perturbation analysing, the up limit of the measuring force error caused by the calibration matrix are computed. The static performance of the sensor is investigated. By sensor’s dynamic modeling, the sensor’s dynamic performance is investigated also. A dynamic compensator is designed, according to the pole assinment. Also the simulation result of the dynamic compensation is presesnted.
This dissertation has made a study of the auto-calibration on the six axis force/torque sensor by two approachs. First, the auto-calibration based on a triaxial accelerometer is presented. With the static application of the triaxial accelerometer, the three angles of the gravity vector and the sensor’s coordinate can be measured simultaneous. According to these three angles, the loads applied by the weights can be calculated automatically. This process alleviates the burden of appling the weights, that can acquire the calibration matrix automatically.The auto-calibration method based on the“Shape from Motion”is prsented, and is applied on the two-axis, three axis, two axis force one torque, and six axis force/torque sensors. Further more the collaborative calibration is also realized by this method. By random appling loads on the multi-axis force sensor, and recording the outputs of the sensor, the pseudoinverse of of the sensor’s calibration matrix and the loads applied on the sensor can be recovered. This method not only improves efficiency of the calibration process, but also avoids the errors introduced by the appling loads, which finally improve the precision of the calibration process.
引文
1 Gao X.H, Jin M.H, Jiang L, et al. The HIT/DLR dexterous hand: Work in progress[A]. Proceedings of IEEE International Conference on Robotics and Automation, 2003, 3: 3164-3168.
2 G. Piller. A Compact Six-degree-of-freedom force/torque sensor for assembly robots[A]. Robot Sensors, 1986, 2:67-74.
3 P.C. Watson, S.H. Drake. Pedestal and wrist force sensors for automatic assembly[A]. Proc.5th Int. Symp. on Industrial Robots, 1975:501-511.
4 Robot Technology. Kogon Page Ltd, London. 1983, 2.
5 B.Shimano. On Force Sensing Information and Its Use in Controlling Manipulators[A]. Proc. of the 8th Int. Symposium on Industrial Robots, 1979.
6 黄心汉, 胡建元, 王健. 一种非径向三梁结构六维腕力传感器弹性体及其优化设计[J]. 机器人, 1992, 14(5): 1-7.
7 H.V Brussel. Force sensing for advanced robot control[A]. Proc.of the 5th Int. Conf. on Robot Vision and Sensory Control. 1980.
8 D. Kerr. Analysis properties and design of a stewart platform transducer[J]. Journal of Mechanisms Transmissions and Automation Design, 1989,25(111).
9 M. Sorli, S. Pastoreli. Six-axis reticulated structure force/torque sensor with adaptable performance[J]. Mechatronics, 1995,5(6):585-601.
10 王洪瑞, 陈贵林, 高峰, 等. 基于 Stewart 平台的 6 维力传感器各向同性的进一步分析[J]. 机械工程学报, 2000, 36(4):49-52.
11 Wang Hongrui, Gao Feng, Huang Zhen. Design of 6-axis force/torque sensor based on stewart platform related to isotropy[J]. Chinese Journal of Mechanical Engineering. 1998(3).
12 Makoto Kaneko. Twin-head six-axis force sensors[J]. IEEE Transactions on Robotics and Automation. 1996,12(1):146-154.
13 B Kuebler, U Seibold, G Hirzinger. Development of actuated and sensor integrated forceps for minimally invasive robotic surgery[A]. Int J Medical Robotics and Computer Assisted Surgery 2005, 1(3):96–107.
14 Seibold U, Kuebler B, Hirzinger G. Prototype of instrument for minimally invasive surgery with 6-axis force sensing capability[A]. Proceedings of the 2005 IEEE International Conference on Robotics and Automation (ICRA),2005.498-503.
15 Seibold U, Hirzinger G. A 6-Axis force/torque sensor design for haptic feedback in minimally invasive robotic surgery[A]. Proceedings of the 2nd VDE World Microtechnologies Congress. 2003.
16 Peter Berkelman, Louis L. Whitcomb, A miniature microsurgical instrument tip force sensor for enhanced force feedback during robot-assisted manipulation[A]. IEEE Transactions on Robotics and Automation, 2003,19(5):917-922.
17 Jan Peirs, Joeri Clijnen, Dominiek Reynaerts. A micro optical force sensor for force feedback during minimally invasive robotic surgery[J]. Sensors and Actuators A, 2004,115:447-455.
18 J.H Kim, D.I Kang, H.H Shin. Design and analysis of a column type multi-component force /moment sensor[J]. Measurement, 2003,33:213–219.
19 Yu-Chung Huang, Ying-Hwi Chang, Wei-Shu Chang. A new structure for a fingertip force sensor[J]. IEEE/ASME Transactions on Mechatronics, 1999,4(3):293-300.
20 Robert D, Lorenz, Karl M. Meyer, et al. A novel compliant four degree-of-freedom robotic fingertip sensor[J]. IEEE Transactions Industry Application 1900,26(4):613-619.
21 Juhyoung Lee, Youngil Youm, Wankyun Chung. The development of POSTECH hand 5[A]. Proceedings of the 2004 IEEE International Conference on Robotics & Automation New Orleans, LA, 2004, April:3386-3390.
22 Multi-Axis Force/Torque Sensor New 2005 Catalog[M]. ATI industrial automation, 2005:3-20.
23 G.Hirzinger, J. Dietrich. Multisensory robots and sensorbased path generation. Robotics and Automation[A]. Proceedings. 1986 IEEE International Conference on, 1986,3(Apr):1992-2001.
24 王全玉, 赵杰, 蔡鹤皋. 六维柔性腕力传感器的设计研究[J]. 机器人, 1996,18(5):279-284.
25 王全玉. 机器人柔性腕力传感器及其作业性能研究[D]. 博士学位论文. 哈尔滨工业大学. 1997.
26 T.B.Martin, R.O.Ambrose, et al. Tactile gloves for autonomous grasping with the NASA/DARPA robonaut[A]. Proceedings of the 2004 IEEE International Conference on Robotics & Automation New Orleans, LA April 2004:1713-1718.
27 Lyndon B. Compact Tactile sensors for robot fingers[DB/OL].www.nasatech.com
28 Shadow Tactile Fingertip[DB/OL]. www.manu-systems.com
29 Taktile Sensorsysteme von Weiss Robotics[DB/OL]. www.weiss-robotics.de
30 Koh Hosoda, Yasunori Tadaa, Minoru Asada. Anthropomorphic robotic soft fingertip with randomly distributed receptors[J]. Robotics and Autonomous Systems, 2006,54:104–109.
31 George Adelman. Encyclopedia of neuroscience[M]. Birkhauser Boston, Inc: 1987.1227-1231.
32 A Cranny, D P J Cotton, P H Chappell, et al. Thick-film force, slip and temperature sensors for a prosthetic hand[J]. Measurement Science and Technology, 2005,16:931-941.
33 Zhongcheng Wu, Ming Meng, Fei Shen. Interaction force measurement of robotic manipulator based on 12DOF force sensor[A]. Information Acquisition, Proceedings International Conference on, 2004:240- 243.
34 吴仲城, 孟明, 申飞. 一体化结构六维加速度传感器设计[J]. 仪器仪表学报, 2004,25(4):302-303.
35 W. L. Jin, C. D. Mote, Jr. Development of a six-component miniature force sensor using silicon micromachining and conventional machining technologies[J]. IEEE Transactions on instrumentation and measurement, 1998,47(3):715-719.
36 W.L.Jin, C.D.Mote, Jr. A six-component silicon micro force sensor[J]. Sensors and Actuators A, 1998,65:109-115.
37 W.L. Jin, C.D.Mote, Jr. Development and calibration of a sub-millimeter three-component force sensor[J]. Sensors and Actuators A, 1998,65:89-94.
38 徐科军, 李成. 多维力传感器静态解耦的研究[J]. 合肥工业大学学报(自然科学版), 1999, 22(2):1-6.
39 徐科军. 传感器动态特性的实用研究方法[M]. 合肥: 中国科学技术大学出版社, 1999.
40 Yusaku Fujii, Hiroyuki Fujimoto. Proposal for an impulse response evaluation method for force transducers[J]. Measurement Science & Technology,1999,10:N31–N33.
41 Yusaku Fujii. Proposal for a step response evaluation method for force transducers[J]. Measurement Science & Technology, 2003,14:1741–1746.
42 Yusaku Fujii. Dynamic Force calibration methods for force transducers[A]. IMTC 2004 - Inslmmentalion and Measurement Technology Conference Corno.Italy, 2004,(3):352-356.
43 Yusaku Fujii. Measurement of steep impulse response of a force transducer[J]. Measurement Science & Technology, 2003,14:65–69.
44 Yusaku Fujii, Koji Ohgushi, Takuro Tojo. A proposal for a dynamic-response-evaluation method for torque transducers[J]. Measurement Science & Technology, 1999, 10:N142–N144.
45 Yusaku Fujii. Amethod for calibrating force transducers against oscillation force[J]. Measurement Science & Technology, 2003,14:1259–1264.
46 徐科军, 朱志能. 六维腕力传感器阶跃响应的实验建模[J]. 机器人, 第 22 卷第 4 期 2000 年 7 月:251-255.
47 刘正士, 陆益民. 基于正交实验设计的多轴腕力传感器静动特性的数值分析[J]. 机械设计, 1999, 2(2):19-24.
48 高理富, 宋宁 等, 航天机器人用六维腕力传感器动态特性研究[J]. 机器人, 2002,24(4):319-323.
49 徐科军 , 唐霆 . 传感器零极点相消动态补偿法 [J]. 科学通报 , 1994, 3927(5):57-66.
50 徐科军, 唐霆. 腕力传感器动态补偿数字滤波器设计[J].自动化仪表, 1994, 15(10):5-7.
51 徐 科 军 , 张 颖 等 . 腕 力 传 感 器 动 态 补 偿 研 究 [J]. 计 量 学 报 , 1997,18(2):116-121.
52 徐科军, 殷铭, 张颖. 腕力传感器的一种动态解耦方法[J]. 应用科学学报, 1999,17(1):39-44.
53 徐科军, 李成. 多维力传感器迭代动态解耦方法[J]. 中国机械工程, 1999,10(1):46-48.
54 徐科军, 殷铭. 腕力传感器动态补偿与解耦方法的研究[J]. 计量学报, 1998,19(4):275-279.
55 Tai-Ran Hsu, MEMS & Microsystems: Design and Manufacture[M] McGraw-Hill Companies, Inc, 2002.
56 Stephen D. Senturia. 微系统设计[M]. 北京:电子工业出版社, 2004.
57 Stephen Beeby, Graham Ensell, Michael Kraft, et al. MEMS Mechanical Sensors[M]. Norwood MA:Artech House, Inc.
58 梅涛. 微传感器与微机器人研究[D]. 博士学位论文. 中国科学技术大学. 2001.
59 陈光华, 邓金祥. 新型电子薄膜材料[M]. 北京:化学工业出版社, 2002.
60 陈宝清. 离子镀及溅射技术[M]. 北京:国防工业出版社,1990.
61 尹福炎. 电阻应变计技术六十年(一)电阻应变计的由来、发展及展望[J]. 传感器世界, 1998, 8:27-32.
62 尹福炎. 薄膜应变片与传感器[J]. 传感器世界, 1997, 6:30-36.
63 Jih-Fen Lei, H A Will. Thin-film thermocouples and strain gauge technologies for engine applications[J]. Sensors and Actuators A, 1998, (65):187-193.
64 Jack Chen, Jonathan Engel, Chang Liu. Development of polymer-based artificial haircell using surface micromachining and 3D assembly[A]. The 12th International Conference on Solid State Sensors, Actuators and Microsystems, Boston, 2003,June 8-12:1035-1038.
65 A. García-Alonso, J. García, et al. Strain sensitivity and temperature influence on sputtered thin films for piezoresistive sensors[J]. Sensors and Actuators A: Physical, 1993,37-38(3-4):784-789.
66 L.I. Belic. AES, AFM and TEM studies of NiCr thin films for capacitive humidity sensors[J]. Thin Solid Films, 1998,317:173-177.
67 唐武, 徐可为. Au/NiCr/Ta 多层金属膜退火后的电阻率异常增大[J]. 金属学报, 2003,2:172-174.
68 唐武, 徐可为. Au/NiCr/Ta 多层金属膜的表面粗糙度和纳米压入硬度的研究[J]. 金属学报, 2002, 5:449-452.
69 唐武, 马幼平 等. Au/NiCr/Ta 和 Au/NiCr 多层金属膜的划痕特征载荷. 金属学报, 2002, 4:404-410.
70 唐武, 徐可为. 微波集成电路(MIC)中 Au/NiCr/Ta 多层金属膜粗糙化机理的AFM 研究[J]. 真空科学与技术, 2003, 2:140-142.
71 于映, 陈跃. NiCr 溅射薄膜内应力的研究[J]. 真空电子技术, 2000,5:9-12.
72 W. Brückner, J. Schumanna, et al. Resistance behaviour and interdiffusion of layered CuNi-NiCr films[J]. Thin Solid Films. 1995,258:236-246.
73 S. Baunack, W. Bruckner, et al. Auger electron spectroscopy study of interdiffusion, oxidation and segregation during thermal treatment of NiCr/CuNi(Mn)/NiCr thin films[J]. Application Surface Sciciece, 1999,216:144-145.
74 姜力. 具有力感知功能的机器人灵巧手手指及控制的研究[D]. 哈尔滨工业大学博士学位论文, 2001.
75 杨磊. 基于指尖传感器的机器人灵巧手阻抗控制[D]. 哈尔滨工业大学博士学位论文. 2004.
76 陶宝琪, 王妮. 电阻应变式传感器[M]. 北京:国防工业出版社, 1993.
77 张如一, 沈观林, 李朝弟. 应变电测与传感器[M]. 北京:清华大学出版社,1999.
78 Imam H.Kazi , P.M.Wild , T.N.Moor , et al. The electromechanical behavior of nichrome (80y20 wt.%) film[J]. Thin Solid Films, 2003,433:337–343.
79 Yong Kwon, Nam-Hoon Kim, Gwang-Pyo Choi, et al. Structural and surface properties of NiCr thin films prepared by DC magnetron sputtering under variation of annealing conditions[J]. Microelectronic Engineering, 2005,82:314–320.
80 尹福炎. 当今传感器用应变计的四大关键工艺[J]. 衡器, 1995,77(2):7-13.
81 陈曦 . 聚对二甲苯在电子领域中应用的新进展 [J]. 电子工艺技术 , 2002,23(4):146-148.
82 李鸣明, 孙燕, 赵宏. 三维扫描技术在薄膜厚度分布测量中的应用[J]. 光子学报, 2004,33(1):90-92.
83 邓元龙, 姚建铨, 阮双琛, 等. 纳米厚度薄膜外差椭偏测量技术的研究[J]. 光学技术, 2005,3(3):391-393.
84 赖倩茜, 李戈扬, 竺品芳. 纳米薄膜厚度的 X 射线测量[J]. 理化检验—物理分册, 2000,36(12):549-551.
85 宋敏, 李波欣, 郑亚茹. 利用光学方法测量薄膜厚度的研究[J]. 光学技术, 2004,30(1): 103-106.
86 David Pekker , Leonid Pekker. A method for determining thickness and optical constants of absorbing thin films[J]. Thin Solid Films, 2003,425:203–209.
87 Michael F. Toney, C. Mathew Mate, et al. Thickness measurements of thin perfluoropolyether polymer films on silicon and amorphous-hydrogenated carbon with x-ray reflectivity, ESCA and optical ellipsometry[J]. Journal of Colloid and Interface Science, 2000,225:219–226.
88 M.C. Baykul , A. Balcioglu. AFM and SEM studies of CdS thin films produced by an ultrasonic spray pyrolysis method[J]. Microelectronic Engineering, 2000,51:703–713.
89 陈雄标,袁哲俊,姚英学. 机器人用六维腕力传感器标定研究[J], 机器人, 1997,19(1):7-11.
90 Antonio Bicchi. A criterion for the optimal design of multi-axis force sensors[A] Memorandum Report Massachusetts Inst. of Tech., Cambridge. Artificial Intelligence Lab 1990,10.
91 陈雄标,袁哲俊,姚英学. 多维力传感器设计的评价准则与优化设计研究[J]. 哈尔滨工业大学学报, 1997,29(4):88-92.
92 许德章, 吴仲城. 机器人六维腕力传感器耦合矩阵的确定与摄动分析[J]. 仪器仪表学报, 2005,26(1):75-81.
93 孙继广. 矩阵扰动分析(第二版)[M]. 北京:科学出版社, 2001.
94 王昊. 基于随动式机器人的航天服关节力学特性测试系统研究[D]. 哈尔滨工业大学博士学位论文, 2004.
95 郑红梅, 刘正士, 王勇. 机器人六维腕力传感器标定方法和标定装置的研究计量学报[J], 2005,26(1):43-45.
96 Richard M. Voyles. Toward gesture-based programming: agent-based haptic skill acquisition and interpretation[D]. CMU-RI-TR-97-36.
97 Conrad J. Poelman, Takeo Kanade A paraperspective factorization method for shape and motion recovery[J]. IEEE Transctions on Pattern Analysis and Machine Intelligence, 1997,19(3):206-218.
98 戴华, 矩阵论[M]. 北京:科学出版社. 2001.
99 曹长修, 自适应消除干扰的新算法[J]. 自动化学报, 1986,12(2):146-154.
100 许光辉, 胡光锐, 宋阳. 基于奇异值分解的非线性滤波算法[J]. 上海交通大学学报, 2005,39(4):626-628.
101 温广瑞,张西宁,屈梁生. 奇异值分解技术在声音信息分离中的应用[J]. 西安交通大学学报, 2003,37(1): 37-40.
102 付京逊, 机器人学控制传感技术视觉智能[M]. 北京:中国科学技术出版社,1989.
