Boris Sokolov

Integrated modelling of agile enterprise networks

Agile enterprise networks, also called as Collaborative Networks (CN) emerge in a decentralised and dynamical way instead of former static hierarchical cooperation and value chains. The CN modelling and optimisation issues differ from those in traditional supply chains and require combined application of various modelling techniques. The paper aims to elaborate an approach Decentralised Integrated Modelling Approach (DIMA) of the CN integrated modelling with the use of elements drawn from different approaches such as control theory, systems theory, operations research and distributed artificial intelligence. The basics of the multi-disciplinary treatment of the CN are developed to contribute to the establishing foundations of the agile networks theory as called for by increasing number of researchers. Our research elaborates a special multi-disciplinary technique for CN modelling to provide flexible application of various modelling frameworks (analytical, simulation and heuristics) as well as their combinations (multiple-model complexes). The results show that the multi-disciplinary treatment of agile production networks allows a comprehensive and realistic planning and control problems formulation and solution. These techniques should allow solving various CN management problems with available data, interlinking with the other relevant problems, etc. as well avoiding the isolated problems solution and incoherent and non-consistent model fragments.

Situational Modelling for Structural Dynamics Control of Industry-Business Processes and Supply Chains

The paper describes a flexible information technology for decision making support by situational modelling of possible alternative structures of industry-business processes and supply chains both in statics and dynamics. It provides a state analysis and prognosis of spatial hierarchical objects. The technology allows to use data coming from simulation modules of the object structural components, embedded geographic information system and expert system equally.

Adaptive control of the top blown oxygen converter process considering temperature and unknown external disturbances

The paper considers the control of the top blown oxygen converter process by exploiting the equations of the process in order to compensate for unknown disturbances. These disturbances are caused by additives dissolving in the converter. Necessity to take into account temperature of reaction has resulted in increase of number of the equations in the in the mathematical model of process. Thus the speed of reaction now depends on the metal temperature. The dissipation region of the controller is evaluated. A nonlinear observer is developed for estimating the metal analysis in real time. It is shown that the observer is asymptotically stable and a region of attraction, where the estimation error converges, is derived. Adaptive control of this process is constructed.

Stability of the solutions of the observation error equations at control of the steel converter process

Consideration was given to the nonlinear observer with the feedback sensitivity function which showed itself well in estimating the carbon and silicon content in the converter melt on the basis of the nonlinear model with time-constant parameters. In the case of controlling the melt temperature by the cooling additives, the plant parameters vary in time, which requires additional studies. For the estimation error under permanent perturbations where the matrix of the linear part is time-dependent, the zero solution of the equation system was proved to be stable.

Relay control of the steel converter process by means of cooling additives

The problem of steel converter control by means of continuously manipulated cooling additives has been considered before. Here a more realistic situation when the control signal takes only two values i. e. zero or highest possible is treated.