Parallel implementation of an adaptive generalized predictive control algorithm

The Adaptive Generalized Predictive Control (AGPC) algorithm can be speeded up using parallel processing. Since the AGPC algorithm needs to be fed with knowledge of the plant transfer function, the parallelization of a standard Recursive Least Squares (RLS) estimator and a GPC predictor is discussed here. Also a parallel implementation of Gauss-Jordan elimination method to handle matrix inversion is presented. The parallel algorithm is implemented over a small network of T805 transputers so that it can afterwards be mapped to an equivalent DSP processor network. Execution times and efficiency results are presented to show the performance of the parallel algorithm.