基于神经网络的进化机器人行为集成方法的研究
|
摘要
以往对机器人的研究大多是在已知的、结构化环境中进行的，研究人员对于机器人的
自身以及机器人的工作环境都有精确的先验知识。这种传统的机器人设计方法必然存在一
些问题，主要表现在：（1）设计者必须要具有机器人及其工作环境的先验知识，即首先要
建立机器人及其工作环境的数学模型。（2）如果对机器人自身以及工作环境知识不断完善
的话，那么就要不断地修改硬件和软件上的设计，给工程实现带来难以评估的工作量。（3）
机器人必须精确地按照环境的内部模型进行规划的结果来运行，适应能力差。
进化机器人是机器人研究领域中的一个重要分支，由于它具有简洁的结构和高度的自
主能力，这一方向正受到越来越多的国内外专家学者的重视。具有代表性的行为主义思想
认为，设计智能机器人的有效途径应象生物体进化那样，采用“自下而上”方式，以感知
-行动作为基础，在与环境的交互中学习。在设计智能机器人过程中，一个关键问题是如
何来实现“行为主义”的思想、在与环境的交互中学习行为动作。
本论文则基于进化神经网络，对机器人行为模型和算法进行了系统研究。研究过程更
加注重机器人的适应性，不注重对环境知识完备性的要求，让机器人就在完全未知环境下
运行。通过机器人自身对环境的感知，来建立环境的模型，并且具有自恢复能力。机器人
通过不断的学习，完善自身的适应能力，依靠与环境不断的交互来获得知识，并通过反复
调整环境模型与自身的模型，最终学会在未知环境中运行。本文主要研究内容及创新成果
包括：
1、基于并行遗传算法和L-系统模拟自然进化和生长学习过程，提出一个人工神经网络结
构设计算法。算法对生成网络的产生式规则进行编码来约束网络的搜索空间，并采用
并行算法降低运算时间。实验证明该算法能有效提高网络结构设计的性能和收敛速度。
2、提出了基于神经网络的进化机器人避碰、趋近和沿壁行为学习算法。文中首先提出了
新的机器人模拟环境和机器人模型，构建了基于神经网络的进化学习系统；然后对具
有进化学习机制的机器人基本行为学习系统进行了仿真实验，并对仿真结果进行分析
与讨论。
3、基于神经网络方法提出了进化机器人避碰、趋近、沿壁行为的高级组合行为切换学习
算法。首先对机器人模拟环境进行了增强设计，给出采用神经网络实现进化学习系统
的方法，然后对具有进化学习机制的机器人组合行为切换学习系统进行了仿真实验，
并对仿真结果进行分析与讨论，最后指出了进一步研究方向。
4、基于神经网络集成了进化多机器人编队行为及避碰、趋近、随机等基本行为，以控制
摘要
机器人规划路径、避开危险并同时保持队形。文中首先在研究了自主机器人构架
（AU RARA）的基础上，利用神经网络实现了机器人高级组合行为的进化学习，然后对
具有进化学习机制的机器人编队行为系统进行了仿真实验，并对多种模拟环境中的不
同编队行为类型进行了性能分析与讨论。
5、提出了一个基于生长神经网络的进化机器人行为算法，新算法的主要特点是：l）、迢
过自然选择对神经同络进行进化，并能自主实 沿人进磁、移动、复制和攻击等行
为；2）、开发了一个自主机器人模拟环境，对所提出的算法进行运行恻试．模拟结果
证明，生长神经冈络系统是研究渐增进化的有效工具，新算法能够有效地实现机召人
创新行为；由于采用了生长系统，这些网络在复制过程中克服了传统进化网络的基因
型相同而表现型不同的固有缺点，系统能在长期渐增进化中处于准．收敛状态．
— —’—’－’”、r－一、——”‘—、一’”—”””’”’—”一’”“—一”“『”’”一’一’——”’————””“’”t
本文所研究的各种算法并不局限于上述应用，对函数优化、组合优化、板式识别、图
象压络等计算机应用领域也有广泛的应用前景．
Traditional robot research was based on known structured environment, and researchers
have accurate knowledge about robot and its environment. The main problems about this method
are: (1) robot designers must have accurate knowledge about robot and its environment; (2) the
hardware or software must be modified if robot and its environment need improving, this
increased the burden of engineering implementation; (3) the adaptive ability of robot is weak
because it must run under the plantation result from internal model of environment.
Evolutionary robot is an important branch of robots science and technology more and more
AX specialists pay attention to it because of its simplified structure and robust autonomous ability.
Behaviorist believed that intelligent robot should be designed by using bottom-up method and
should learn behaviors based on sensing-reactive by interacting with the environment.
In this thesis, we present some new models and algorithms of robot behavior by using
evolutionary neural network. The new model takes more emphasis on adaptive property of the
robot, and let the robot run under the unknown environment. The robot can build the model of
world by sensing environment and have self recovery ability. By learning continually and
interacting with the environment, the robot can keep on adjusting the world and its own model
and can finally run the environment. Our creative work include the following parts:
1. A new method for designing artificial neural network architectures, which is based on
system and genetic algorithm, is presented in this paper. Production rules and parallel
algorithm are used to solve traditional neural network design problems. The experiment
results have proved that the algorithm can improve network performance and the speed of
converge.
2. Obstacle avoidance, target approach, and wall following learning of evolutionary robot are
realized by using artificial neural network in this paper. First, a robot learning environment
and a robot model are presented and the implementation of evolutionary learning system is
discussed. Then, the simulation experiments for basic behaviors and switch learning system
of intelligent robot that adopted evolutionary learning mechanism are carried out. Finally,
the simulation results are analyzed.
3. The behavior switch learning of evolutionary robot is realized by using artificial neural
network in this paper. A robot learning environment and a robot model are presented and the
implementation of evolutionary learning system is discussed. The simulation experiments
1~(E)
are can-ied out for switch learning system of intelligent robot that adopted evolutionary
learning mechanism. Finally, the simulation results are analyzed and the future research
direction is given.
4. Obstacle avoidance, target approach, random and formation learning of evolutionary robot
are realized by using artificial neural network in this paper. First, an advanced robot
combined behavior learning method is presented and the implementation of evolutionary
learning system is discussed by using neural network and AuRA architecture. Then, the
simulation experiments are carried out for formation learning system of intelligent robot that
adopted evolutionary learning mechanism. Finally, the simulation results of several type of
formation type are given and analyzed.
5. This
自身以及机器人的工作环境都有精确的先验知识。这种传统的机器人设计方法必然存在一
些问题，主要表现在：（1）设计者必须要具有机器人及其工作环境的先验知识，即首先要
建立机器人及其工作环境的数学模型。（2）如果对机器人自身以及工作环境知识不断完善
的话，那么就要不断地修改硬件和软件上的设计，给工程实现带来难以评估的工作量。（3）
机器人必须精确地按照环境的内部模型进行规划的结果来运行，适应能力差。
进化机器人是机器人研究领域中的一个重要分支，由于它具有简洁的结构和高度的自
主能力，这一方向正受到越来越多的国内外专家学者的重视。具有代表性的行为主义思想
认为，设计智能机器人的有效途径应象生物体进化那样，采用“自下而上”方式，以感知
-行动作为基础，在与环境的交互中学习。在设计智能机器人过程中，一个关键问题是如
何来实现“行为主义”的思想、在与环境的交互中学习行为动作。
本论文则基于进化神经网络，对机器人行为模型和算法进行了系统研究。研究过程更
加注重机器人的适应性，不注重对环境知识完备性的要求，让机器人就在完全未知环境下
运行。通过机器人自身对环境的感知，来建立环境的模型，并且具有自恢复能力。机器人
通过不断的学习，完善自身的适应能力，依靠与环境不断的交互来获得知识，并通过反复
调整环境模型与自身的模型，最终学会在未知环境中运行。本文主要研究内容及创新成果
包括：
1、基于并行遗传算法和L-系统模拟自然进化和生长学习过程，提出一个人工神经网络结
构设计算法。算法对生成网络的产生式规则进行编码来约束网络的搜索空间，并采用
并行算法降低运算时间。实验证明该算法能有效提高网络结构设计的性能和收敛速度。
2、提出了基于神经网络的进化机器人避碰、趋近和沿壁行为学习算法。文中首先提出了
新的机器人模拟环境和机器人模型，构建了基于神经网络的进化学习系统；然后对具
有进化学习机制的机器人基本行为学习系统进行了仿真实验，并对仿真结果进行分析
与讨论。
3、基于神经网络方法提出了进化机器人避碰、趋近、沿壁行为的高级组合行为切换学习
算法。首先对机器人模拟环境进行了增强设计，给出采用神经网络实现进化学习系统
的方法，然后对具有进化学习机制的机器人组合行为切换学习系统进行了仿真实验，
并对仿真结果进行分析与讨论，最后指出了进一步研究方向。
4、基于神经网络集成了进化多机器人编队行为及避碰、趋近、随机等基本行为，以控制
摘要
机器人规划路径、避开危险并同时保持队形。文中首先在研究了自主机器人构架
（AU RARA）的基础上，利用神经网络实现了机器人高级组合行为的进化学习，然后对
具有进化学习机制的机器人编队行为系统进行了仿真实验，并对多种模拟环境中的不
同编队行为类型进行了性能分析与讨论。
5、提出了一个基于生长神经网络的进化机器人行为算法，新算法的主要特点是：l）、迢
过自然选择对神经同络进行进化，并能自主实 沿人进磁、移动、复制和攻击等行
为；2）、开发了一个自主机器人模拟环境，对所提出的算法进行运行恻试．模拟结果
证明，生长神经冈络系统是研究渐增进化的有效工具，新算法能够有效地实现机召人
创新行为；由于采用了生长系统，这些网络在复制过程中克服了传统进化网络的基因
型相同而表现型不同的固有缺点，系统能在长期渐增进化中处于准．收敛状态．
— —’—’－’”、r－一、——”‘—、一’”—”””’”’—”一’”“—一”“『”’”一’一’——”’————””“’”t
本文所研究的各种算法并不局限于上述应用，对函数优化、组合优化、板式识别、图
象压络等计算机应用领域也有广泛的应用前景．
Traditional robot research was based on known structured environment, and researchers
have accurate knowledge about robot and its environment. The main problems about this method
are: (1) robot designers must have accurate knowledge about robot and its environment; (2) the
hardware or software must be modified if robot and its environment need improving, this
increased the burden of engineering implementation; (3) the adaptive ability of robot is weak
because it must run under the plantation result from internal model of environment.
Evolutionary robot is an important branch of robots science and technology more and more
AX specialists pay attention to it because of its simplified structure and robust autonomous ability.
Behaviorist believed that intelligent robot should be designed by using bottom-up method and
should learn behaviors based on sensing-reactive by interacting with the environment.
In this thesis, we present some new models and algorithms of robot behavior by using
evolutionary neural network. The new model takes more emphasis on adaptive property of the
robot, and let the robot run under the unknown environment. The robot can build the model of
world by sensing environment and have self recovery ability. By learning continually and
interacting with the environment, the robot can keep on adjusting the world and its own model
and can finally run the environment. Our creative work include the following parts:
1. A new method for designing artificial neural network architectures, which is based on
system and genetic algorithm, is presented in this paper. Production rules and parallel
algorithm are used to solve traditional neural network design problems. The experiment
results have proved that the algorithm can improve network performance and the speed of
converge.
2. Obstacle avoidance, target approach, and wall following learning of evolutionary robot are
realized by using artificial neural network in this paper. First, a robot learning environment
and a robot model are presented and the implementation of evolutionary learning system is
discussed. Then, the simulation experiments for basic behaviors and switch learning system
of intelligent robot that adopted evolutionary learning mechanism are carried out. Finally,
the simulation results are analyzed.
3. The behavior switch learning of evolutionary robot is realized by using artificial neural
network in this paper. A robot learning environment and a robot model are presented and the
implementation of evolutionary learning system is discussed. The simulation experiments
1~(E)
are can-ied out for switch learning system of intelligent robot that adopted evolutionary
learning mechanism. Finally, the simulation results are analyzed and the future research
direction is given.
4. Obstacle avoidance, target approach, random and formation learning of evolutionary robot
are realized by using artificial neural network in this paper. First, an advanced robot
combined behavior learning method is presented and the implementation of evolutionary
learning system is discussed by using neural network and AuRA architecture. Then, the
simulation experiments are carried out for formation learning system of intelligent robot that
adopted evolutionary learning mechanism. Finally, the simulation results of several type of
formation type are given and analyzed.
5. This
引文
[1] Brooks, R.A. "A Robust Layered Control System for a Mobile Robot", IEEE Journal of Robotics and Automation, RA-2, pp 14-23． 1996
[2] Brooks, R.A. "A Robot that Walks: Emergent Behaviors from a Carefully Evolved Network", MIT AI memo 1091． 1989
[3] Gomi, T. "Non-Cartesian Robotics", Robotics and Autonomous System 18 pp 169-184, Elsevier. 1996
[4] Harvey, F. "Evolutionary Robotics and SAGA": the case for hill crawling and tournament selection", in Artificial Life 3 Proceedings, C.Langton, editor, Santa Fe Institute Studies in the Sciences of Complexity, Proc. Vol XVI, pp 299-326． 1997
[5] Floreano, D., and Mondada, F. "Automatic Creation of an Autonomous Agent:Genetic Evolution of Neural Network Driven Robot", in D.Cliff, P.Husbands, J.Meyer and S.W. Wilson, editors, From Animals to Animats Ⅲ: Proceedings of the Third International Conference on Simulation of Adaptive Behavior, MIT Press-Bradford Brooks, Cambridge, MA. 1994
[6] Gomi, T., and Griffith, A. "Evolutionay Robotics-An Overview", invited paper at International Conference on Evolutionary Computation (ICEC'96) Nagoya, Japan. 1996
[7] Gomi, T. "Subsumption Robots and the Application of Intelligent Robots to the Service Industry", Internal report, Applied AI Systems, Inc. 1992
[8] Nishi, G., Shinbori, S., Gomi, T., Ide, K., Maheral, P., "Mechanimals for Mechquarium", AAI Report. 1995
[9] Tsuda, I. "Chaotic View of the Brain", K.K. Science. 1990
[10] Gomi, T. And Laurence, J.C. "Behavior-based AI Techniques for Vehicle Control", Vehicle Navigation & Information Systems Conference(VNIS'93) , Ottawa, Canada. 1993
[11] Mondada, F., Franzi, E, lenne, P. "Mobile Robot Miniaturisation: A Tool for Investigation in Control Algorithms", in Third International Symposium on Experimental Robotics, Kyoto, Japan. 1993
[12] Brooks, R.A., "A Robot that Walks: Emergent Behavior from a Carefully Evolved Network", Neural Computation 1:2, 1997
[13] Brooks, R.A., "Artifical Life and Real Robots", European Conference on Artificial Life(ECAL'91) , Paris, France, December 1991
[14] Davis L., "Handbook of Genetic Algorithm", Van Nostrand Reinhold Publisher, New York, 1991
[15] Floreano, D. and Mondada, F., "Autonumous and Self-Sufficient: Emergent Homing Behaviors in a Mobile Robot", LAMI Technical Report No R94． 14I
[16] Floreano, D. and Mondada, F., "Active Perception, navigation, Homing, and Grasping: An Autonomous Perspective", In Proceedings, From Perception to Action, J.D. Nicoud, and Ph. Gaussier(Eds), IEEE Press, Los Alamitos, CA, 1994
[17] Gomi, T., and Ulvr, J., "Artifical Emotions as Emergent Phenomena", 2~(nd) IEEE International Workshop on Robot and Human Communication(RO-MAN'93) , Tokyo, Japan, November 1993
[18] Harvey, I., Husbands, P., Cliff, D., "Genetic Convergence in a Species of Evolved Robot Control Architectures", University of Sussex CSRP 267, Brighton, U.K., February 1993
[19] Maes, P., "A Spreading Activation Network for Action Selection", Intelligent Autonomous System-2(IAS-2) , Amsterdam, December 1989
[20] Maes, P., "Learning Behavior Networks from Experience", First European Conference on Artificial Life ECAL(91) , MIT Press, Paris, France, December 1991
[21] Mataric, M., "A Disributed Model for Mobile Robot Environment-Learning and Navigation", MIT Artificial Intelligence Laboratory Technical Report AI-TR-1228
[22] Nolfi, S., Floreano, D., Miglino, O., Mondada, F., "How to Evolve Autonomous Robots: Different Approaches in Evolutionary Robotics", Fourth International Workshop on the Synthesis and Simulation of Living Systems, Artificial Life IV, Boston, MA, July, 1994
[1] Rumelhart D E, Hinton G E. Williams R.J. Learning internal representations by error propagation, Parallel Distributed Processing: Explorations in Microstructures of Cognition, 1986,l(l):318-362
[2] Tanese R. Distributed genetic algorithms. In: Schaffer J D ed, Proceedings of the Third International Conference on Genetic Algorithms. San Mateo, CA: Morgan Kaufmann. 1989, 434-439
[3] Dodd N. Optimization of network structure using genetic algorithms. In: Widrow B, Angeniol B ods, Proceedings of the International Neural Network Conferonce, INNC-90-Paris. Kluwer, Dordrecht. 1990, 693-696
[4] Harik G, Cantu-Paz E, Goldberg D E, Miller B. The gambler's ruin problem, genetic algorithms, and the sizing of populations. In: Back T ed, Proceedings of the Fourth International Conference on Evolutionary Computation. New York: IEEE Press. 1997, 7-12
[5] Jacobs R A, Jordan M I, Barto A G. Task decomposition through competition in a modular conneetionist Architecture: the what and where vision tasks. In: Cognitive Science. 1991(15), 219-250
[6] Mandelbrot B B. The fractal geometry of nature. Freeman, San Francisco. 1982.
[7] Cantu-Paz E. Designing efficient master-slave parallel genentic algorithms(IlliGAL Report No.97004). Urbana, IL: Unniversity of Illinois at Urbana-Champaign. 1997
[8] Posner M I, Peterson S E, Fox P T et al. Localization of cognitive operations in the human brain. In: Science. 1988(240), 1627-1631
[9] Prusinkiewicz P, Hanan J. Lindenmayer systems, fractals and plants. Springer-Verlag, New York. 1989.
[10] Whitley D, Starkweather T. Genitor Ⅱ: A distributed genetic algorithm. Journal of Experimental and Theoretical Artificial Intelligence. 1990
[11] Solla S A. Learning and generalization in layered neural networks: the contiguity problem. In: Personnas L, Dreyfus G eds, Neural networks: from models to applications. I.D.S.E.T, Paris. 1989(10), 168-177
[12] Murre J M J. Categorization and learning in neural networks. Modelling and implementation in a modular framework. Dissertation, Leidcn University. 1992.
[13] 王洪燕，杨敬安．并行遗传算法研究进展．计算机科学，26(6)，1999
[14] 王洪燕，杨敬安．并行进化BP神经网络．合肥工业大学学报，22(3)，1999
[15] 王洪燕，杨敬安．关于主从式分层并行遗传算法的研究，第五届全国计算机应用联合学术会议论文集，1999，北京
[1] Rumelhart D.E.,Hinton G.E. Williams R.J. Learning internal representations by error propagation, Parallel Distributed Processing: Explorations in Microstructures of Cognition, 1986,1(1) .318-362
[2] Holland J H. Adaptation in natural and artificial systems. Ann Arbor, MI: University of Michigan Press, 1975． 162-169
[3] Mitchell A.Potter, Kenneth A. De Jong. Evolving Neural Networks with collaborative Species. Proceedings of the 1995 Summer Computer Simulation Conference, 1995,7,24-26
[4] Grosso P. B. Computer simulations of genetic adaptation: Parallel subcomponent interaction in a multilocus model. Doctoral dissertation, The University of Michigan. 1985
[5] Cohoon J. P., Hegde S. U., Martin W. N., Richards D. Punctuated equilibria: A parallel genetic algorithm. In Grefenstette, J. J. (Ed.), Proceedings of the Second International Conference on Genetic Algorithms. 1987:148-154
[6] Tanese R. Distributed genetic algorithms. In Schaffer, J. D. (Ed.), Proceedings of the Third International Conference on Genetic Algorithms,San Mateo, CA: Morgan Kaufmann. 1989:434--439．
[7] Whitley D., Starkweather T. Genitor II: A distributed genetic algorithm. Journal of Experimental and Theoretical Artificial Intelligence, 1990
[8] Hills D. Co-evolving parasites improce simulated evolution as an optimization procedure. In C.Langtion, C.Tayor,J.Farmer, and S.Rasmussen(Eds.),Artificial Life Ⅱ, 1984． 313-324
[9] Husbands P. And F.Mill. Simulated coevolution sa the mechanism for emergent planning and scheduliing. In R.Belew and L. Brooker(Eds.), Proceedings of the Fourth International Conference on Genetic Algorithms, 1991:264-270
[10] David E. Moriarty,Risto Miikkulainen. Evolving Neural Networks to Focus Minimax Search. Proceedings of the Twelfth National Conference on Artificial Intelligence,1994．
[11] Mitchell A. Potter, Kenneth A. De Jong ,John J. Grefenstette. A Coevolutionary Approach to Learning Sequential Decision Rules. Proceedings of the Sixth International Conference(ICGAA95) , 1995, 7: 15-19
[12] Talbi E.-G., Bessiere, P. A parallel genetic algorithm for the graph partitioning problem. In Proc. Of the International Conference on Supercomputing. 1991, 6
[1] R Brooks. "Intelligence without reason". in Proc. Int. Joint Conf. Artificial Intell., 1993
[2] R Brooks and A Flynn. "A robust layered control system for a mobile robot". IEEE Trans. Robotics Automat., vol. 2, no. 1, 1986
[3] Hee Rak Beom, Hyung Suck Cho. "A Sensor-based Navigation for a Mobile Robot Using Fuzzy Logic and Reinforcement Learning". IEEE Transaction On System, Man, and Cybernetics. 1995, 25(3)
[4] Andrew G Barto, Richard S. Sutton, Peter S. Brouwer. "Associative Search Network: A Reinforcement Learning Associative Memory". Biol. Cybern, 1981, 40: 201-202
[5] D. E. Goldberg. "Genetic Algorithms in Search, Optimization, and Machine Learning". Addison-Wesley, Reading, MA, 1989．
[6] S. J. Louis and G. J. E. Rawlins. "Designer genetic algorithms: Genetic algorithms in structure design." In Proceedings of the Fourth International Conference on Genetic Algorithms, pages 53-60． Morgan Kauffman, San Mateo, CA, 1991．
[7] A. Murray and S. J. Louis. "Adapting control strategies for situated autonomous agents." In Proceedings of the Florida Artificial Intelligence Research Symposium, 1995．
[8] A. Murray and S. J. Louis, "Design strategies for evolutionary robotics." In E. A. Yfantis, editor, Proceedings of the Third Golden West International Conference on Intelligent Systems, Pages 609-616． Kluwer Academic Press, 1995．
[9] Jingan Yang, A Probabilistic Model for Dynamic Motion Planning in Partially Known Environments Based on Discrete Events, In Proc of 3rd IFAC Symposium on Intelligent Autonomous Vehicles, Madrid, Spain, 1998． 3 and reprinted by Automatica, 1999．
[10] Yang Jingan, A Neural Paradigm for Time-Varying Motion Segmentation, Journal of Computer Science & Technology, 14 (6) : 1-13, 1999． 11．
[l]R.C.Arkin, "Motor schema based mobile robot navigation", Int. J. Robot. Res., vol. 8, no. 4, pp. 92-112, 1989．
[2] R.C. Arkin and T. R. Balch, "AuRa: Principles and practice in review", J. Exper. Theor. Artif. Intell., vol. 9, no. 2, 1997．
[3] U. S. Army, Field Manual No 7-7J, Washington, DC, 1986．
[4] D. C. Brogan and K. K. Hodgins, "Group behaviors for systems with significant dynamics". Auton. Robots, vol. 4, no. 1, pp. 137-153, Mar. 1997．
[5] R. Brooks, "A robust layered control system for a mobile robot", IEEE J. Robot. Automat., vol. RA-2, p. 14, Feb. 1986．
[6] Q.Chen and J. Y. S. Luh, "Coordiantion and control of a group of small mobile robots", in Proc. 1994 IEEE Int. Conf. Robot. Automat., San Diego, CA, 1994, pp. 2315-2320．
[7] D. J. Cook, P. Gmytrasiewicz, and L. B. Holder, "Decision-theoretic cooperative sensor planning", IEEE Trans. Pattern Anal. Machine Intell., vol. 18, pp. 1013-1023, Oct. 1996．
[8] J. M. Cullen, E. Shaw, and H. A. Baldwin, "Methods for measuring the three-dimencional structure of fish schools", Animal Beh., vol. 13, pp. 534-543, 1965．
[9] U. S. Air Force, Air Combat Command Manual 3-3, Washington, DC, 1992．
[10] D. W. Gage, "Command control for many-robot systems", Unmanned Syst. Mag., vol. 10, no. 4, pp. 28-34, 1992．
[11] D. Longer, J. Rosenblatt, and M. Hebert, "A behavior-based system for off-road navigation", IEEE Trans. Robot. Automat., vol. 10, pp. 776-783, Dec. 1994．
[12] D. MacKenzie, R. Arkin, and J. Cameron, "Multiagent mission specificaiton and execution", Auton. Robots, vol. 4, no. 1, pp. 29-52, 1997．
[13] M. Mataric, "Designing emergent behaviors: From local interactions to collective intelligence", in Proc. Int. Conf. Simulationi of Adaptive Behavior: From Animals to Animats 2, 1992, pp. 432-441．
[14] M. Mataric, "Minimizing complexity in controlling a mobile robot population", in Proc. 1992 IEEE Int. Conf. Robot. Automat., Nice, France, May 1992, pp. 830-835．
[15] L. Parker, "Designing control laws for cooperative agent teams", in Proc. 1993 IEEE Int. Conf. Robot. Automat., 1993, pp. 582-587
[16] L. E. Parker, Heterogeneous Multi-Robot Cooperation, Ph.D. dissertation, Dept. Electr. Eng. Comput. Sci., Mass. Inst. Of Technol., Cambridge, MA, 1994．
[17] C. Reynolds, "Flocks, herds and schools: A distributed behavioral model", Comput. Graph., vol. 21, no. 4, pp. 25-34, 1987．
[18] J. Rosenblatt, "DAMN: A distributed architecture for mobile navigation", in Working Notes AAAI 1995 Spring Symp. Lessons Learned for Implemented Software Architectures for Physical Agents, Palo Alto, CA, Mar. 1995．
[19] X. Tu and D. Terzopoulos, "Artificial fished: Physics, locomotion, perception, behavior", in Proc. SIGGRAPH 94 Conf., Orlando, FL. July 1994, pp. 43-50．
[20] S. L. Veherencamp, "Individual, kin, and group selection", in Handbook of Behavioral Neurobiology, Volume 3: Social Behavior and Communication, P. Marler and J. G. Vandenvergh, Eds. New York: Plenum, 1987, pp. 354-382．
[21] P. K. C. Wang, "Navigation strategies for multiple autonomous robots moving in formation", J. Robot. Syst., vol. 8, no. 2, pp. 177-195, 1991．
[22] H. Yamaguchi, "Adaptive formation control for distributed autonomous mobile robot groups", in Proc. 1997 IEEE Conf. Robot. Automat., Albuquerque, NM, Apr. 1997．
[23] Jingan Yang, A Probabilistic Model for Dynamic Motion Planning in Partially Known Environments Based on Discrete Events, In Proc of 3~(rd) IFAC Symposium on Intelligent Autonomous Vehicles, Madrid, Spain, 1998． 3 and reprinted by Automatica, 1999．
[24] Yang Jingan, A Neural Paradigm for Time-Varying Motion Segmentation, Journal of Computer Science & Technology, 14 (6) : 1-13, 1999． 11．
[1] Zaera N., Cliff D. and Bruten J. Evolving collective behaviours in synthetic fish. In Maes P. , Mataric M., Meyer J. A. eds., Proceedings of SAB96, 635-644． MIT Press Bradford Books. 1996
[2] Ray T. S. An approach to the synthesis of life. In Langton, C.; Taylor, C.; Farmer, J.; and Rasmussen, S., eds., Artificial Life II, 371--408． Redwood City, CA: Addison-Wesley. 1991
[3] Yaeger L. Computational genetics, physiology, metabolism, neural systems, learning, vision, and behavior or polyworld: Life in a new context. In Langton, C. G., ed., Artificial Life Ⅲ, 263-298． 1993
[4] Channon A. and Damper R. Perpetuating evolutionary emergence. In Proceedings of SAB98． MIT Press. 1998
[5] Harvey I. Evolutionary robotics and SAGA: the case for hill crawling and tournament selection. In Langton, C. G., ed., Artificial Life Ⅲ. 1993
[6] Koza J. R. Genetic Programming. Cambridge, MA: MIT Press/ Bradford Books. 1992
[7] Gruau F. Artificial cellular development in optimization and compilation. Technical report, Psychology department, Stanford Universit Ursityalo Alto, CA. 1996
[8] Lindenmayer A. Mathematical models for cellular interaction in development. Journal of Theoretical Biology 18:280-315． Parts I and Ⅱ. 1968
[9] Kitano H. Designing neural networks using genetic algorithms with graph generation system. Complex Systems 4:461--476． 1990
[10] Boers E. J. and Kuiper H. Biological metaphors and the design of modular artificial neural networks. Master's thesis, Departments of Computer Science and Experimental Psychology, Leiden University. 1992
[2] Brooks, R.A. "A Robot that Walks: Emergent Behaviors from a Carefully Evolved Network", MIT AI memo 1091． 1989
[3] Gomi, T. "Non-Cartesian Robotics", Robotics and Autonomous System 18 pp 169-184, Elsevier. 1996
[4] Harvey, F. "Evolutionary Robotics and SAGA": the case for hill crawling and tournament selection", in Artificial Life 3 Proceedings, C.Langton, editor, Santa Fe Institute Studies in the Sciences of Complexity, Proc. Vol XVI, pp 299-326． 1997
[5] Floreano, D., and Mondada, F. "Automatic Creation of an Autonomous Agent:Genetic Evolution of Neural Network Driven Robot", in D.Cliff, P.Husbands, J.Meyer and S.W. Wilson, editors, From Animals to Animats Ⅲ: Proceedings of the Third International Conference on Simulation of Adaptive Behavior, MIT Press-Bradford Brooks, Cambridge, MA. 1994
[6] Gomi, T., and Griffith, A. "Evolutionay Robotics-An Overview", invited paper at International Conference on Evolutionary Computation (ICEC'96) Nagoya, Japan. 1996
[7] Gomi, T. "Subsumption Robots and the Application of Intelligent Robots to the Service Industry", Internal report, Applied AI Systems, Inc. 1992
[8] Nishi, G., Shinbori, S., Gomi, T., Ide, K., Maheral, P., "Mechanimals for Mechquarium", AAI Report. 1995
[9] Tsuda, I. "Chaotic View of the Brain", K.K. Science. 1990
[10] Gomi, T. And Laurence, J.C. "Behavior-based AI Techniques for Vehicle Control", Vehicle Navigation & Information Systems Conference(VNIS'93) , Ottawa, Canada. 1993
[11] Mondada, F., Franzi, E, lenne, P. "Mobile Robot Miniaturisation: A Tool for Investigation in Control Algorithms", in Third International Symposium on Experimental Robotics, Kyoto, Japan. 1993
[12] Brooks, R.A., "A Robot that Walks: Emergent Behavior from a Carefully Evolved Network", Neural Computation 1:2, 1997
[13] Brooks, R.A., "Artifical Life and Real Robots", European Conference on Artificial Life(ECAL'91) , Paris, France, December 1991
[14] Davis L., "Handbook of Genetic Algorithm", Van Nostrand Reinhold Publisher, New York, 1991
[15] Floreano, D. and Mondada, F., "Autonumous and Self-Sufficient: Emergent Homing Behaviors in a Mobile Robot", LAMI Technical Report No R94． 14I
[16] Floreano, D. and Mondada, F., "Active Perception, navigation, Homing, and Grasping: An Autonomous Perspective", In Proceedings, From Perception to Action, J.D. Nicoud, and Ph. Gaussier(Eds), IEEE Press, Los Alamitos, CA, 1994
[17] Gomi, T., and Ulvr, J., "Artifical Emotions as Emergent Phenomena", 2~(nd) IEEE International Workshop on Robot and Human Communication(RO-MAN'93) , Tokyo, Japan, November 1993
[18] Harvey, I., Husbands, P., Cliff, D., "Genetic Convergence in a Species of Evolved Robot Control Architectures", University of Sussex CSRP 267, Brighton, U.K., February 1993
[19] Maes, P., "A Spreading Activation Network for Action Selection", Intelligent Autonomous System-2(IAS-2) , Amsterdam, December 1989
[20] Maes, P., "Learning Behavior Networks from Experience", First European Conference on Artificial Life ECAL(91) , MIT Press, Paris, France, December 1991
[21] Mataric, M., "A Disributed Model for Mobile Robot Environment-Learning and Navigation", MIT Artificial Intelligence Laboratory Technical Report AI-TR-1228
[22] Nolfi, S., Floreano, D., Miglino, O., Mondada, F., "How to Evolve Autonomous Robots: Different Approaches in Evolutionary Robotics", Fourth International Workshop on the Synthesis and Simulation of Living Systems, Artificial Life IV, Boston, MA, July, 1994
[1] Rumelhart D E, Hinton G E. Williams R.J. Learning internal representations by error propagation, Parallel Distributed Processing: Explorations in Microstructures of Cognition, 1986,l(l):318-362
[2] Tanese R. Distributed genetic algorithms. In: Schaffer J D ed, Proceedings of the Third International Conference on Genetic Algorithms. San Mateo, CA: Morgan Kaufmann. 1989, 434-439
[3] Dodd N. Optimization of network structure using genetic algorithms. In: Widrow B, Angeniol B ods, Proceedings of the International Neural Network Conferonce, INNC-90-Paris. Kluwer, Dordrecht. 1990, 693-696
[4] Harik G, Cantu-Paz E, Goldberg D E, Miller B. The gambler's ruin problem, genetic algorithms, and the sizing of populations. In: Back T ed, Proceedings of the Fourth International Conference on Evolutionary Computation. New York: IEEE Press. 1997, 7-12
[5] Jacobs R A, Jordan M I, Barto A G. Task decomposition through competition in a modular conneetionist Architecture: the what and where vision tasks. In: Cognitive Science. 1991(15), 219-250
[6] Mandelbrot B B. The fractal geometry of nature. Freeman, San Francisco. 1982.
[7] Cantu-Paz E. Designing efficient master-slave parallel genentic algorithms(IlliGAL Report No.97004). Urbana, IL: Unniversity of Illinois at Urbana-Champaign. 1997
[8] Posner M I, Peterson S E, Fox P T et al. Localization of cognitive operations in the human brain. In: Science. 1988(240), 1627-1631
[9] Prusinkiewicz P, Hanan J. Lindenmayer systems, fractals and plants. Springer-Verlag, New York. 1989.
[10] Whitley D, Starkweather T. Genitor Ⅱ: A distributed genetic algorithm. Journal of Experimental and Theoretical Artificial Intelligence. 1990
[11] Solla S A. Learning and generalization in layered neural networks: the contiguity problem. In: Personnas L, Dreyfus G eds, Neural networks: from models to applications. I.D.S.E.T, Paris. 1989(10), 168-177
[12] Murre J M J. Categorization and learning in neural networks. Modelling and implementation in a modular framework. Dissertation, Leidcn University. 1992.
[13] 王洪燕，杨敬安．并行遗传算法研究进展．计算机科学，26(6)，1999
[14] 王洪燕，杨敬安．并行进化BP神经网络．合肥工业大学学报，22(3)，1999
[15] 王洪燕，杨敬安．关于主从式分层并行遗传算法的研究，第五届全国计算机应用联合学术会议论文集，1999，北京
[1] Rumelhart D.E.,Hinton G.E. Williams R.J. Learning internal representations by error propagation, Parallel Distributed Processing: Explorations in Microstructures of Cognition, 1986,1(1) .318-362
[2] Holland J H. Adaptation in natural and artificial systems. Ann Arbor, MI: University of Michigan Press, 1975． 162-169
[3] Mitchell A.Potter, Kenneth A. De Jong. Evolving Neural Networks with collaborative Species. Proceedings of the 1995 Summer Computer Simulation Conference, 1995,7,24-26
[4] Grosso P. B. Computer simulations of genetic adaptation: Parallel subcomponent interaction in a multilocus model. Doctoral dissertation, The University of Michigan. 1985
[5] Cohoon J. P., Hegde S. U., Martin W. N., Richards D. Punctuated equilibria: A parallel genetic algorithm. In Grefenstette, J. J. (Ed.), Proceedings of the Second International Conference on Genetic Algorithms. 1987:148-154
[6] Tanese R. Distributed genetic algorithms. In Schaffer, J. D. (Ed.), Proceedings of the Third International Conference on Genetic Algorithms,San Mateo, CA: Morgan Kaufmann. 1989:434--439．
[7] Whitley D., Starkweather T. Genitor II: A distributed genetic algorithm. Journal of Experimental and Theoretical Artificial Intelligence, 1990
[8] Hills D. Co-evolving parasites improce simulated evolution as an optimization procedure. In C.Langtion, C.Tayor,J.Farmer, and S.Rasmussen(Eds.),Artificial Life Ⅱ, 1984． 313-324
[9] Husbands P. And F.Mill. Simulated coevolution sa the mechanism for emergent planning and scheduliing. In R.Belew and L. Brooker(Eds.), Proceedings of the Fourth International Conference on Genetic Algorithms, 1991:264-270
[10] David E. Moriarty,Risto Miikkulainen. Evolving Neural Networks to Focus Minimax Search. Proceedings of the Twelfth National Conference on Artificial Intelligence,1994．
[11] Mitchell A. Potter, Kenneth A. De Jong ,John J. Grefenstette. A Coevolutionary Approach to Learning Sequential Decision Rules. Proceedings of the Sixth International Conference(ICGAA95) , 1995, 7: 15-19
[12] Talbi E.-G., Bessiere, P. A parallel genetic algorithm for the graph partitioning problem. In Proc. Of the International Conference on Supercomputing. 1991, 6
[1] R Brooks. "Intelligence without reason". in Proc. Int. Joint Conf. Artificial Intell., 1993
[2] R Brooks and A Flynn. "A robust layered control system for a mobile robot". IEEE Trans. Robotics Automat., vol. 2, no. 1, 1986
[3] Hee Rak Beom, Hyung Suck Cho. "A Sensor-based Navigation for a Mobile Robot Using Fuzzy Logic and Reinforcement Learning". IEEE Transaction On System, Man, and Cybernetics. 1995, 25(3)
[4] Andrew G Barto, Richard S. Sutton, Peter S. Brouwer. "Associative Search Network: A Reinforcement Learning Associative Memory". Biol. Cybern, 1981, 40: 201-202
[5] D. E. Goldberg. "Genetic Algorithms in Search, Optimization, and Machine Learning". Addison-Wesley, Reading, MA, 1989．
[6] S. J. Louis and G. J. E. Rawlins. "Designer genetic algorithms: Genetic algorithms in structure design." In Proceedings of the Fourth International Conference on Genetic Algorithms, pages 53-60． Morgan Kauffman, San Mateo, CA, 1991．
[7] A. Murray and S. J. Louis. "Adapting control strategies for situated autonomous agents." In Proceedings of the Florida Artificial Intelligence Research Symposium, 1995．
[8] A. Murray and S. J. Louis, "Design strategies for evolutionary robotics." In E. A. Yfantis, editor, Proceedings of the Third Golden West International Conference on Intelligent Systems, Pages 609-616． Kluwer Academic Press, 1995．
[9] Jingan Yang, A Probabilistic Model for Dynamic Motion Planning in Partially Known Environments Based on Discrete Events, In Proc of 3rd IFAC Symposium on Intelligent Autonomous Vehicles, Madrid, Spain, 1998． 3 and reprinted by Automatica, 1999．
[10] Yang Jingan, A Neural Paradigm for Time-Varying Motion Segmentation, Journal of Computer Science & Technology, 14 (6) : 1-13, 1999． 11．
[l]R.C.Arkin, "Motor schema based mobile robot navigation", Int. J. Robot. Res., vol. 8, no. 4, pp. 92-112, 1989．
[2] R.C. Arkin and T. R. Balch, "AuRa: Principles and practice in review", J. Exper. Theor. Artif. Intell., vol. 9, no. 2, 1997．
[3] U. S. Army, Field Manual No 7-7J, Washington, DC, 1986．
[4] D. C. Brogan and K. K. Hodgins, "Group behaviors for systems with significant dynamics". Auton. Robots, vol. 4, no. 1, pp. 137-153, Mar. 1997．
[5] R. Brooks, "A robust layered control system for a mobile robot", IEEE J. Robot. Automat., vol. RA-2, p. 14, Feb. 1986．
[6] Q.Chen and J. Y. S. Luh, "Coordiantion and control of a group of small mobile robots", in Proc. 1994 IEEE Int. Conf. Robot. Automat., San Diego, CA, 1994, pp. 2315-2320．
[7] D. J. Cook, P. Gmytrasiewicz, and L. B. Holder, "Decision-theoretic cooperative sensor planning", IEEE Trans. Pattern Anal. Machine Intell., vol. 18, pp. 1013-1023, Oct. 1996．
[8] J. M. Cullen, E. Shaw, and H. A. Baldwin, "Methods for measuring the three-dimencional structure of fish schools", Animal Beh., vol. 13, pp. 534-543, 1965．
[9] U. S. Air Force, Air Combat Command Manual 3-3, Washington, DC, 1992．
[10] D. W. Gage, "Command control for many-robot systems", Unmanned Syst. Mag., vol. 10, no. 4, pp. 28-34, 1992．
[11] D. Longer, J. Rosenblatt, and M. Hebert, "A behavior-based system for off-road navigation", IEEE Trans. Robot. Automat., vol. 10, pp. 776-783, Dec. 1994．
[12] D. MacKenzie, R. Arkin, and J. Cameron, "Multiagent mission specificaiton and execution", Auton. Robots, vol. 4, no. 1, pp. 29-52, 1997．
[13] M. Mataric, "Designing emergent behaviors: From local interactions to collective intelligence", in Proc. Int. Conf. Simulationi of Adaptive Behavior: From Animals to Animats 2, 1992, pp. 432-441．
[14] M. Mataric, "Minimizing complexity in controlling a mobile robot population", in Proc. 1992 IEEE Int. Conf. Robot. Automat., Nice, France, May 1992, pp. 830-835．
[15] L. Parker, "Designing control laws for cooperative agent teams", in Proc. 1993 IEEE Int. Conf. Robot. Automat., 1993, pp. 582-587
[16] L. E. Parker, Heterogeneous Multi-Robot Cooperation, Ph.D. dissertation, Dept. Electr. Eng. Comput. Sci., Mass. Inst. Of Technol., Cambridge, MA, 1994．
[17] C. Reynolds, "Flocks, herds and schools: A distributed behavioral model", Comput. Graph., vol. 21, no. 4, pp. 25-34, 1987．
[18] J. Rosenblatt, "DAMN: A distributed architecture for mobile navigation", in Working Notes AAAI 1995 Spring Symp. Lessons Learned for Implemented Software Architectures for Physical Agents, Palo Alto, CA, Mar. 1995．
[19] X. Tu and D. Terzopoulos, "Artificial fished: Physics, locomotion, perception, behavior", in Proc. SIGGRAPH 94 Conf., Orlando, FL. July 1994, pp. 43-50．
[20] S. L. Veherencamp, "Individual, kin, and group selection", in Handbook of Behavioral Neurobiology, Volume 3: Social Behavior and Communication, P. Marler and J. G. Vandenvergh, Eds. New York: Plenum, 1987, pp. 354-382．
[21] P. K. C. Wang, "Navigation strategies for multiple autonomous robots moving in formation", J. Robot. Syst., vol. 8, no. 2, pp. 177-195, 1991．
[22] H. Yamaguchi, "Adaptive formation control for distributed autonomous mobile robot groups", in Proc. 1997 IEEE Conf. Robot. Automat., Albuquerque, NM, Apr. 1997．
[23] Jingan Yang, A Probabilistic Model for Dynamic Motion Planning in Partially Known Environments Based on Discrete Events, In Proc of 3~(rd) IFAC Symposium on Intelligent Autonomous Vehicles, Madrid, Spain, 1998． 3 and reprinted by Automatica, 1999．
[24] Yang Jingan, A Neural Paradigm for Time-Varying Motion Segmentation, Journal of Computer Science & Technology, 14 (6) : 1-13, 1999． 11．
[1] Zaera N., Cliff D. and Bruten J. Evolving collective behaviours in synthetic fish. In Maes P. , Mataric M., Meyer J. A. eds., Proceedings of SAB96, 635-644． MIT Press Bradford Books. 1996
[2] Ray T. S. An approach to the synthesis of life. In Langton, C.; Taylor, C.; Farmer, J.; and Rasmussen, S., eds., Artificial Life II, 371--408． Redwood City, CA: Addison-Wesley. 1991
[3] Yaeger L. Computational genetics, physiology, metabolism, neural systems, learning, vision, and behavior or polyworld: Life in a new context. In Langton, C. G., ed., Artificial Life Ⅲ, 263-298． 1993
[4] Channon A. and Damper R. Perpetuating evolutionary emergence. In Proceedings of SAB98． MIT Press. 1998
[5] Harvey I. Evolutionary robotics and SAGA: the case for hill crawling and tournament selection. In Langton, C. G., ed., Artificial Life Ⅲ. 1993
[6] Koza J. R. Genetic Programming. Cambridge, MA: MIT Press/ Bradford Books. 1992
[7] Gruau F. Artificial cellular development in optimization and compilation. Technical report, Psychology department, Stanford Universit Ursityalo Alto, CA. 1996
[8] Lindenmayer A. Mathematical models for cellular interaction in development. Journal of Theoretical Biology 18:280-315． Parts I and Ⅱ. 1968
[9] Kitano H. Designing neural networks using genetic algorithms with graph generation system. Complex Systems 4:461--476． 1990
[10] Boers E. J. and Kuiper H. Biological metaphors and the design of modular artificial neural networks. Master's thesis, Departments of Computer Science and Experimental Psychology, Leiden University. 1992