Dr Mark P. Balenzuela
As technology becomes more prominent and integrated within human life, and the level of automation increases, so do the potential risks of a malfunction. With a successful and reliable fault diagnosis system, autonomous robots can reconfigure themselves during a fault to ensure safe operation continues.
BAYESIAN estimator development
Mark’s research is in part focused on developing accurate and efficient algorithms for Bayesian estimation of both Gaussian mixture model (GMM) and Jump Markovian systems. These estimators can be used to solve fault diagnosis and parameter estimation problems.
System identification is an important real-world problem, which is required to be solved in order to implement model-based control schemes. Mark’s research in parameter estimation includes developing new methodologies in which the expectation-maximisation (EM) algorithm can be applied with minimal computational cost and with improved convergence.