Behavioral task processing for cognitive robots using artificial emotions Evren Daglarli, Hakan Temeltas à , Murat Yesiloglu Istanbul Technical University, Electrical-Electronics Engineering Faculty, Control Engineering Department, 34469 Maslak, Istanbul, Turkey article info Available online 18 April 2009 Keywords: Intelligent robotics Autonomous systems Artificial cognitive models Emotion-based control abstract This paper presents an artificial emotional-cognitive system-based autonomous robot control architecture for a four-wheel driven and four-wheel steered mobile robot. Discrete stochastic state- space mathematical model is considered for behavioral and emotional transition processes of the autonomous mobile robot in the dynamic realistic environment. The term of cognitive mechanism system which is composed from rule base and reinforcement self-learning algorithm explain all of the deliberative events such as learning, reasoning and memory (rule spaces) of the autonomous mobile robot. The artificial cognitive model of autonomous robot control architecture has a dynamic associative memory including behavioral transition rules which are able to be learned for achieving multi-objective robot tasks. Motivation module of architecture has been considered as behavioral gain effect generator for achieving multi-objective robot tasks. According to emotional and behavioral state transition probabilities, artificial emotions determine sequences of behaviors for long-term action planning. Also reinforcement self-learning and reasoning ability of artificial cognitive model and motivational gain effects of proposed architecture can be observed on the executing behavioral sequences during simulation. The posture and speed of the robot and the configurations, speeds and torques of the wheels and all deliberative and cognitive events can be observed from the simulation plant and virtual reality viewer. This study constitutes basis for the multi-goal robot tasks and artificial emotions and cognitive mechanism-based behavior generation experiments on a real mobile robot. & 2009 Elsevier B.V. All rights reserved. 1. Introduction The new generation control architectures of autonomous mobile robots have been designed to be inspired from cognitive mechanism of the human brain. The emotional activations and cognitive events of brain play important role for decision making and long-term deliberative process planning of humans [1]. According to viewpoint of autonomous systems and intelligent robotics, artificial emotions can be considered as trigger of behavioral action sequences [2]. In the autonomous robot control systems, emotional transitions and behavior selection process should be based on probabilistic statistical modeling [3]. The term ‘‘cognitive mechanism model of the architecture’’ explains us all of the deliberative events such as learning, reasoning and memory (rule spaces) of the mobile robot [4]. This system employs a hybrid learning algorithm-based computational model which involves linguistic reasoning (dynamic fuzzy cognitive maps) and stochastic behavioral selection. Indeed, emotions and cogni- tion are indispensable features of human brain for defining the behavioral characteristics. In alives, motivation effect of emotions is thought that define intense of the executing behaviors and activate them by neuropsychologists and cognitive scientists [2]. Motives determine intense of performing task processes of the behaviors during required optimal time frame. According to goals and needs of the mobile robot, this idea supports that a certain behavioral gain coefficient is applied to each behavior of the sequence as adaptive [4]. In literature, emotion–motivation for intentional selection and configuration of behavior-producing modules (EMIB) robot con- trol and computational agent architecture were presented as good examples to emotion and motivation-based robot control archi- tecture in 2002 [2]. This architecture was made of three main levels; behavior system, recommendation level and motivation level. Emotion-based robot control architecture was introduced in 2003 and named as ALEC (asynchronous learning by emotion and cognition) architecture [5] which utilizes hidden Markov model- based stochastic model. Another work on emotion-based struc- ture was presented by Buss [3] in 2004. Advancement in cognitive sciences and control engineering will keep shedding light into this area of research. Problems on decision making analysis of multi-objective robot tasks are one of the biggest challenges for the intelligent robot controller design [5]. In order to reach optimal situations (satisfying the goals), intelligent robots have to overcome more complex tasks (behaviors) which may conflict with each other [6] In the other architectures, while the robots are facing dynamic realistic multi-objective environments, they experience too many decision making problems including solution of ranking process of ARTICLE IN PRESS Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/neucom Neurocomputing 0925-2312/$ - see front matter & 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.neucom.2008.07.018 à Corresponding author. Tel.: +90 2122856701; fax: +90 212 2856700. E-mail address: temeltas@elk.idu.edu.tr (H. Temeltas). Neurocomputing 72 (2009) 2835–2844