Bulletin of the Polish Academy of Sciences Technical Sciences

The main focus of the paper is on the asymptotic behaviour of linear discrete-time positive systems. Emphasis is on highlighting the relationship between asymptotic stability and the structure of the system, and to expose the relationship between null-controllability and asymptotic stability. Results are presented for both time-invariant and time-variant systems.

A suitable use of software packages for optimization problems can give the possibility to formulate design problems of robotic mechanical systems by taking into account the several aspects and behaviours for optimum solutions both in design and operation. However, an important issue that can be even critical to obtain practical solutions can be recognized in a proper identification and formulation of criteria for optimability purposes and numerical convergence feasibility. In this paper, we have reported experiences that have been developed at LARM in Cassino by referring to the abovementioned issues of determining a design procedure for manipulators both of serial and parallel architectures. The optimality criteria are focused on the well-recognized main aspects of workspace, singularity, and stiffness. Computational aspects are discussed to ensure numerical convergence to solutions that can be also of practical applications. In particular, optimality criteria and computational aspects have been elaborated by taking into account the peculiarity and constraint of each other. The general concepts and formulations are illustrated by referring to specific numerical examples with satisfactory results.

In this paper we present results of systematic and comprehensive simulation analysis of the Tsao & Safonov unfalsified controller for complex robot manipulators. In particular, we show that the controller falsification procedure yields the closedloop unfalsified controller, which accomplishes the control objective, within a finite and relatively short time interval with the number of invocations of linear programming based unfalsified controller selection procedure being relatively small. We also present some conclusions resulting from the investigation of the effect of such elements as manipulator structure complexity, prior knowledge about disturbances, reference trajectory and assigned closed-loop spectrum on unfalsified controller performance and computational complexity.

Main topic of the paper is a problem of designing the input-output decoupling controllers for nonholonomic mobile manipulators. We propose a selection of output functions in much more general form than in [1,2]. Regularity conditions guaranteeing the existence of the input-output decoupling control law are presented. Theoretical considerations are illustrated with simulations for mobile manipulator consisting of RTR robotic arm mounted atop of a unicycle which moves in 3D-space.

In this paper we propose a sensor-based navigation method for navigation of wheeled mobile robot, based on the Kohonen self-organising map (SOM). We discuss a sensor-based approach to path design and control of wheeled mobile robot in an unknown 2-D environment with static obstacles. A strategy of reactive navigation is developed including two main behaviours: a reaching the middle of a collision-free space behaviour, and a goal-seeking behaviour. Each low-level behaviour has been designed at design stage and then fused to determine a proper actions acting on the environment at running stage. The combiner can fuse low-level behaviours so that the mobile robot can go for the goal position without colliding with obstacles one for the convex obstacles and one for the concave ones. The combiner is a softswitch, based on the idea of artificial potential fields, that chooses more then one action to be active with diRerent degrees at each time step. The output of the navigation level is fed into a neural tracking controller that takes into account the dynamics of the mobile robot. The purpose of the neural controller is to generate the commands for the servo-systems of the robot so it may choose its way to its goal autonomously, while reacting in real-time to unexpected events. Computer simulation has been conducted to illustrate the performance of the proposed solution by a series of experiments on the emulator of wheeled mobile robot Pioneer-2DX.

In order to understand commands given through voice by an operator, user or any human, a robot needs to focus on a single source, to acquire a clear speech sample and to recognize it. A two-step approach to the deconvolution of speech and sound mixtures in the time-domain is proposed. At first, we apply a deconvolution procedure, constrained in the sense, that the de-mixing matrix has fixed diagonal values without non-zero delay parameters. We derive an adaptive rule for the modification of the de-convolution matrix. Hence, the individual outputs extracted in the first step are eventually still self-convolved. This corruption we try to eliminate by a de-correlation process independently for every individual output channel.

Programming frameworks [1] are application generators with the following components: library of software modules (building blocks out of which the system is constructed), a method for designing new modules that can be appended to the above mentioned library, a pattern according to which ready modules can be assembled into a complete system jointly exerting control over it and realizing the task at hand. The presented transition function based formalism can be applied to specifying programming frameworks for robot controllers executing very diverse tasks. The paper deals with systems consisting of multiple embodied agents, influencing the environment through effectors, gathering information from the environment through sensors and communicating with other agents through communication channels. The presented code patterns pertain to behavioural agents. The formalism was instrumental in the design of MRROC++ robot programming framework, which has been used for producing controllers of single and two manipulator systems performing diverse tasks. The formalism introduces rigor into the discussion of the structure of embodied agent controllers. It is used as the means for the specification of the functions of the components of the control system and the structure of the communication links between them. This structures the implementation of a programming framework, and that in turn makes the coding of specific controllers much easier, both from the point of view of dealing with the hardware configuration of the system and the specific task that has to be executed.

Reasoning with limited computational resources (such as time or memory) is an important problem, in particular in knowledge-intensive embedded systems. Classical logic is usually considered inappropriate for this purpose as no guarantees regarding deadlines can be made. One of the more interesting approaches to address this problem is built around the concept of

active logics. Although a step in the right direction, active logics are just a preliminary attempt towards finding an acceptable solution. Our work is based on the assumption that labelled deductive systems (LDSs) offer appropriate metamathematical methodology to study the problem. As a first step, we have reformulated a pair of active logics systems, namely the memory model and its formalized simplification, the step logic, as LDSs. This paper presents our motivation behind this project, followed by an overview of the investigations on meta-reasoning relevant to this work, and introduces in some reasonable detail the MM system.

The project called “Polish Cardiosurgical Robot” has been developed by Foundation for Cardiac Surgery Development since year 2000. Within the project the telemanipulator to perform the endoscopic cardiosurgical operations has been designed, manufactured and examined. In the following paper the development of construction of arms for RobIn Heart 0, RobIn Heart 1, RobIn Heart 3 versions of the robot as well as the fixing system has been presented. In the preliminary phase of the project the requirements for mechanical construction were analyzed. Additional requirements enhancing functionality of the construction were also defined. Analyses of the planned development of the construction and ways of its possible applications were performed.

Neutralisation of the terrorist explosive devices is a risky task. Such tasks may be carried out by robots in order to protect human life. The article describes chosen design problems concerning the new neutralisation and assisting robot SMR-100 Expert. The robot was to be designed for the use in confined spaces, particularly inside the air-crafts, buses and rail cars. In order to achieve this ambitious plan, new advanced technological designing tools had to be applied. A number of interesting design issues were approached. The successful development of the prototype robot Expert in Poland resulted in the creation of the first intervention robot in the world able to perform all necessary anti-terrorist tasks inside the passenger planes.