Kaveh Hassani

Multi-objective design of state feedback controllers using reinforced quantum-behaved particle swarm optimization

A multi-objective design approach for tuning LQR controllers is proposed.A new reinforced quantum-behaved PSO algorithm is introduced.Comparative studies among nine various techniques are conducted. A novel and generic multi-objective design paradigm is proposed which utilizes quantum-behaved PSO (QPSO) for deciding the optimal configuration of the LQR controller for a given problem considering a set of competing objectives. There are three main contributions introduced in this paper as follows. (1) The standard QPSO algorithm is reinforced with an informed initialization scheme based on the simulated annealing algorithm and Gaussian neighborhood selection mechanism. (2) It is also augmented with a local search strategy which integrates the advantages of memetic algorithm into conventional QPSO. (3) An aggregated dynamic weighting criterion is introduced that dynamically combines the soft and hard constraints with control objectives to provide the designer with a set of Pareto optimal solutions and lets her to decide the target solution based on practical preferences. The proposed method is compared against a gradient-based method, seven meta-heuristics, and the trial-and-error method on two control benchmarks using sensitivity analysis and full factorial parameter selection and the results are validated using one-tailed T-test. The experimental results suggest that the proposed method outperforms opponent methods in terms of controller effort, measures associated with transient response and criteria related to steady-state.

An incremental framework for classification of EEG signals using quantum particle swarm optimization

Classification of electroencephalographic (EEG) signals is a sophisticated task that determines the accuracy of thought pattern recognition performed by computer-brain interface (BCI) which, in turn, determines the degree of naturalness of the interaction provided by that system. However, classifying the EEG signals is not a trivial task due to their non-stationary characteristics. In this paper, we introduce and utilize incremental quantum particle swarm optimization (IQPSO) algorithm for incremental classification of EEG data stream. IQPSO builds the classification model as a set of explicit rules which benefits from semantic symbolic knowledge representation and enhanced comprehensibility. We compared the performance of IQPSO against ten other classifiers on two EEG datasets. The results suggest that IQPSO outperforms other classifiers in terms of classification accuracy, precision and recall.

Visualizing Natural Language Descriptions: A Survey

A natural language interface exploits the conceptual simplicity and naturalness of the language to create a high-level user-friendly communication channel between humans and machines. One of the promising applications of such interfaces is generating visual interpretations of semantic content of a given natural language that can be then visualized either as a static scene or a dynamic animation. This survey discusses requirements and challenges of developing such systems and reports 26 graphical systems that exploit natural language interfaces and addresses both artificial intelligence and visualization aspects. This work serves as a frame of reference to researchers and to enable further advances in the field.

Adaptive animation generation using web content mining

Creating 3D animation is a labor-intensive and time-consuming process requiring designers to learn and utilize a complex combination of menus, dialog boxes, buttons and manipulation interfaces for a given stand-alone animation design software. On the other hand, conceptual simplicity and naturalness of visualizing imaginations from lingual descriptions motivates researchers for developing automatic animation generation systems using natural language interfaces. In this research, we introduce an interactive and adaptive animation generation system that utilizes data-driven techniques to extract the required common-sense and domain-specific knowledge from web. This system is capable of creating 3D animation based on users lingual commands. It uses the user interactions as a relevance feedback to learn the implicit design knowledge, correct the extracted knowledge, and manipulate the dynamics of the virtual world in an active and incremental manner. Moreover, system is designed based on a multi-agent methodology which provides it with distributed processing capabilities and cross-platform characteristics. In this paper, we will focus on information retrieval agent which is responsible for extracting numeric data utilized in object attributes, spatiotemporal relations, and environment dynamics using web mining techniques.

A software-in-the-loop simulation of an intelligent microsatellite within a virtual environment

Rapid growth in space missions necessitates the onboard intelligence, which creates autonomous space systems by providing high level decision making, robust execution of decisions, and automatic fault repairing. Mostly, autonomous space systems are implemented as hybrid architectures with a few conceptual layers. Validating the stability and evaluating the performance of an autonomous architecture is critical for space missions. Software-in-the-loop simulation is a suitable approach for addressing this demand. However, the data acquired from simulation is represented as alphanumeric values or diagrams, which needs to be interpreted. In this paper, we propose an intelligent architecture to provide onboard autonomy for an observation micro-satellite. The architecture integrates the low level physical actions with conceptual decision making ability in a hierarchical manner. To evaluate the proposed architecture, we have implemented a distributed software-in-the-loop simulation to simulate the space, satellite, ground stations, and intelligent onboard software. Moreover, for the first time, we have used virtual reality to visualize the satellites autonomous behavior in the orbit. It lets the users have a high level feedback from integrated simulation. Scenario-based evaluations have shown the stability and efficiency of the proposed architecture.

A case study on collective intelligence based on energy flow

In this paper, we propose a stochastic scheme for modeling a multi-species prey-predator artificial ecosystem in order to investigate the influence of energy flow on ecosystem lifetime and stability. Inhabitants of this environment are a few species of herbivore and carnivore birds. In this model, collective behavior emerges in terms of flocking, breeding, competing, resting, hunting, escaping, seeking, and foraging. Ecosystem is defined as a combination of prey and predator species with inter-competition among species within the same level of the food chain, and intra-competition among those belonging to different levels of the food chain. Some energy variables are also introduced as functions of behaviors to model the energy within the ecosystem. Experimental results of 11,000 simulations analyzed by Cox univariate analysis and hazard function suggest that only five corresponding energy variables out of eight aforementioned behaviors influence the ecosystem lifetime. Also, results of survival analysis show that among pairwise interactions between energy factors, only two interactions affect the system lifetime, including interaction between flocking and seeking energies, and interaction between flocking and hunting energies. These results match the observations of real life birds, which use flocking behavior for flexible movements, efficient foraging, social learning, and reducing predation risks.

Simulating collective intelligence of bio-inspired competing agents

Modeling a stochastic multi-species prey-predator artificial ecosystem with two levels of food chain.Investigating the influence of energy flow on the ecosystem lifetime.Defining energy variables as functions of behaviors.Analyzing model on 11,000 simulations by a Cox univariate analysis and a hazard function.Results match the behaviors of real ecosystems. In this paper, a comprehensive model is introduced to investigate the influence of energy flow on the lifetime of stochastic multi-species prey-predator artificial ecosystems. The model consists of a non-stationary hosting environment with food sources and a few species of competing herbivore and carnivore birds that can perform several individual and collective behaviors such as flocking, breeding, competing, resting, hunting, escaping, seeking, and foraging. The experimental results of 11,000 simulations analyzed by Cox univariate analysis and hazard function suggest that only a fraction of associated energy variables and pairwise interactions between them influence the lifetime. The proposed stochastic model can be utilized to simulate the complex multi-agent systems and their emergent behavior. Also, the proposed statistical analysis of energy flow to estimate the system stability and lifetime can be generalized to other physical and economical complex systems.

An experimental study on semi-invasive acupuncture-based EEG signal acquisition

EEG is the most frequently applied method for capturing the brain activity due to its high temporal resolution and portability, and its low cost and health risks. However, EEG signals have very low signal to noise ratio due to the effects of scalp, skull, and many other layers as well as noise generated by physiological and non-physiological artifacts. Furthermore, preparation of EEG monitoring equipment and making proper contact between skin and electrodes is a tedious and time-consuming task due to presence of hair and different skull shapes. In this paper, we report the experimental attempts on improving the accuracy of EEG acquisition using a semi-invasive approach which utilizes acupuncture-based needle penetration to alleviate the effect of sculpt on the EEG signals and enhance the preparation efficiency. High level cluster analysis and low level signal analysis on real-life data recorded for nine physical, lingual, and motor imagery tasks suggest that contrary to our expectations, the proposed method is not effective.

An Intelligent Architecture for Autonomous Virtual Agents Inspired by Onboard Autonomy

Intelligent virtual agents function in dynamic, uncertain, and uncontrolled environments, and animating them is a chaotic and error-prone task which demands high-level behavioral controllers to be able to adapt to failures at lower levels of the system. On the other hand, the conditions in which space robotic systems such as spacecraft and rovers operate, inspire by necessity, the development of robust and adaptive control software. In this paper, we propose a generic architecture for developing autonomous virtual agents that let them to illustrate robust deliberative and reactive behaviors, concurrently. This architecture is inspired by onboard autonomous frameworks utilized in interplanetary missions. The proposed architecture is implemented within a discrete-event simulated world to evaluate its deliberative and reactive behaviors. Evaluation results suggest that the architecture supports both behaviors, consistently.

On designing migrating agents: from autonomous virtual agents to intelligent robotic systems

In the realm of multi-agent systems, migration refers to the ability of an agent to transfer itself from one embodiment such as a graphical avatar into different embodiments such as a robotic android. Embodied agents usually function in a dynamic, uncertain, and uncontrolled environment, and exploiting them is a chaotic and error-prone task which demands high-level behavioral controllers to be able to adapt to failure at lower levels of the system. The conditions in which space robotic systems such as spacecraft and rovers operate, inspire by necessity, the development of robust and adaptive control software. In this paper, we propose a generic architecture for migrating and autonomous agents inspired by onboard autonomy which enables the developers to tailor the agents embodiment by defining a set of feasible actions and perceptions associated with the new body. Evaluation results suggest that the architecture supports migration by performing consistent deliberative and reactive behaviors.