Research On Artificial Fish’s Intelligent Behavioral Control

Artificial life is a product mixed with multiple subjects like information and lifescience, which is one of the most popular research fields in21century. Usually its studyobjects are natural life and main research contents are the simulations of mechanism ofnatural life processes. Its ultimate research goal is extending natural life characters.Through the simulation of natural life, it not only can help us comprehend nature andbenefit from it better, but it also can help us research related domains like naturalanimation and intelligent robot, so the research on artificial life has an imponderablesignificance.At present, virtual natural biological based on software is an important researchbranch of artificial life, namely, by modeling the behavior of natural biological andusing computers or other equipment to simulate. But now there are still someshortcomings in simulating the behaviors of artificial fish:(1)At present, the decision ofindividual artificial fish’s behavioral are only according to its external environmentinformation. Neglecting its information of inner state and tacitly approve its inner statein ideal state.(2)Some researchers have proved by experiments that the behaviors ofcreatures are closely related with their own emotional state. Now the behaviors ofindividual artificial fish in models are all neglected this important characteristic whichwill make the behaviors of artificial fish lack certain rationality.(3) The method ofchoosing interactive individuals has a great influence on thebehavior of creature cluster.But now many models are modeling only according to the appearance of creatures’activities and lacking corresponding theoretical on designing interactive individualmechanism.(4)Nowadays many scholars do a simulated display to the cluster ofcreatures’behaviors in one-dimensional scene, which certainly will neglect some relatedproblems such as the turns and collision of creatures.Because fish is representative in nature and its behavior patterns are similar to othernatural creatures, this paper used the artificial fish established by Dr Tu Xiaoyuan as research object, depended on life science and other related disciplines’ researchachievements as theoretical foundations, depended on theories and methods ofbiological intelligence and artificial life as research measures, this paper, combinedartificial life with biological intelligence reasonably, studied the behavioral controlmethod of artificial fish. At last, the paper established artificial fish intelligent behaviorsystem which is fit with natural ecological roles. The paper’s work mainly containsthese aspects:(1) Research on individual artificial fish’s behavior control methodBased on fish had emotional features and emotion would influence bio-behavioraldecision-making ability, this paper introduced emotional factors into the traditionalartificial fish behavioral framework and proposed a individual artificial fish behaviorcontrol model based on emotion. Firstly, the fuzzy reasoning model of emotion wasestablished on the basis of artificial emotion theory. Secondly, in order to carry outbehavior decision, a hybrid model which combined the fuzzy reasoning model ofemotion and CMAC (Cerebellar Model Articulation Controller) neural network waspresented. Finally, according to behavior decision, an artificial fish path selectingstrategy which based on weight was put forward.(2) Research on group artificial fish’s behavior control methodAimed at biological cluster behavior model’s defect in designing interactiveindividual mechanism, this paper proposed a biological cluster behavior model based ontopological distance. At first, based on Ballerini’s research achievement as theoreticalfoundation, adopt a method based on topological distance to choose interactiveindividuals and limited the number of interactive individuals. Then the model portioneda synchronous zone for individuals. At last, aimed at penetration phenomenon whensimulating cluster behavior, considered about system overhead and applicationscenarios, presented a collision processing scheme.(3) Design and realization of artificial fish intelligent behavior simulating animationin virtual environmentBased on related knowledge on computer graphics as theoretical basis, the paperdesigned and realized artificial fish intelligent behavior animation system, usingOpenGL graphics library and VC++. This system simulated individual and groupbehavior control mode of artificial fish, mainly contained predation, escape, cluster andso on. This paper respectively propose an artificial fish’s behavior control model based onemotional and a biological cluster behavior model based on topological distance. Thecontrol method of the artificial fish which has been mentioned in this paper moreconforms to natural characteristics by the theoretical analysis. The simulationexperiment proves that the method is feasible and effective, which provides new idealsto research the behaviors of artificial creatures.