Our objective is the comfortable interaction between the end user and our system. That is why we decided to use X-Windows environment and graphical user interface.
We support a simple and user-friendly way to initialize the whole system, to control the genetic algorithm in runtime and to present its results. We have designed these facilities for the X-Windows environment, but we support also a console-based user input to the system.
The architecture of the whole user interface is component-based. Each part of it (e.g. a control panel, a kernel info receiver, or a parameter editor) is represented by a standalone application (called component). These components communicate with a data server built in the system kernel via PVM messaging services. We have chosen this component-based architecture because of its flexibility and variability. It makes possible to add new components easily and to control the system in the user-friendly environment. Nevertheless, it is not necessary to run these components every time and all the time of the computation. This can apparently reduce the CPU load.
The text-file journal is maintained automatically by the kernel to provide complete review of the system activity even if components are active.
So far we support following components:
The system kernel supports three different interfaces. The input interface is defined by the supported parameters, specified in the parameter file used for the kernel initialization. The format of the parameter file represents a common and easily extensible input interface of the system kernel. The control interface gives full control over the computation of the genetic algorithm. It enables to start, to stop or to pause the computation. This interface enables to load initialization data. The data interface supports the data exchange between the components. Each component can register (and unregister) its requests for statistical data. The system kernel will automatically send all requested data to the component's queue when these data are available. This miscellaneous statistical data are collected and computed during the computation of the genetic algorithm by the special mechanism, which allows to collect and compute statistical data. This module provides all required calculations and necessary data processing. These data are available for displaying (through data interface) and other usage wherever needed.