Intelligent Control of Diesel Generators Using Gain-Scheduling Based on Online External-Load Estimation

The development of an intelligent control solution for a wide range of diesel generators is discussed. Compared with most existing solutions, the advantages of the proposed solution lie in two folds: (i) The proposed control has the plug-and-play capability which is reflected by an automatic recognition procedure when it is plugged into a specific diesel generator, such that some extensive manual-tuning of the installed controller can be significantly reduced; (ii) The proposed control has an real-time adaptability by using the online external load estimation, such that the integrated system can keep a consistent performance for a wide range of operating conditions. Technically, a general nonlinear dynamic model is firstly developed based on fundamental principles of diesel generators. Then, the system parameters of this model can be identified experimentally or partially retrieved from the data-sheet for a specific unit. By combining an online external-load estimation with this specific units model, finally an intelligent control using the online gain scheduling strategy is proposed. The proposed solution is verified and analyzed based on a lab-sized emulator of a diesel generator, where a controlled AC-motor is employed to emulate a diesel engine. The testing results clearly show that the proposed control solution can lead to a better overall system performance than most existing solutions do, especially subject to widely diverse operating conditions.