The development of novel, practice-oriented and reliable instrumentation and control strategies for wastewater treatment plants in order to improve energy efficiency, while guaranteeing process stability and maintenance of high cleaning capacity, has become a priority for WWTP operators due to increasing treatment costs. To achieve these ambitious and even contradictory objectives, this thesis investigates a combination of online measurement systems, computational intelligence and machine learning methods as well as dynamic simulation models. Introducing the state-of-the-art in the fields of WWTP operation, process monitoring and control, three novel computational intelligence enabled instrumentation, control and automation (ICA) methods are developed and presented. Furthermore, their potential for practical implementation is assessed. The methods are, on the one hand, the automated calibration of a simulation model for the Rospe WWTP that provides a basis for the development and evaluation of the subsequent methods, and on the other hand, the development of soft sensors for the WWTP inflow which estimate the crucial process variables COD and NH4-N, and the estimation of WWTP operating states using Self- Organising Maps (SOM) that are used to determine the optimal control parameters for each state. These collectively, provide the basis for achieving comprehensive WWTP optimization. Results show that energy consumption and cleaning capacity can be improved by more than 50%.