This thesis presents an asymptotic approximated analytical model for the performance of interconnection networks in large multiprocessor systems. The accuracy of the model is compared with simulation results and found to be very good for a large parameter, the size of the multiprocessor system. The memory contention problem in large multiprocessor system is sloved by the random arbiter. It is shown that the systems with the random arbiter have a far better performance than the system with a general arbiter.
There has been continuous effort to achieve total automation of the power plants. But due to complexity, nonlinearities and time-varying properties of the system, no success has been reproted if conventional and/or modern control theories are applied. In spite of this, the start-up and shutdown operations are successfully performed by skilled human operators who are able to utilize a great wealth of knowledge and past experiences.