Relevance vector machines based modelling and optimisation for collaborative control parameter design: a case study

摘要

A new collaborative control parameter design strategy is proposed for economic plant control process. The relevance vector machines (RVMs) and genetic algorithms (GAs) are combined to generate the optimal control index table for controllers. More specifically, the probabilistic model based on RVMs is utilised to describe the non-linear behaviours according to the experimental dataset. The evolution-based optimisation model based on GAs is used for collaborative design of the optimum control parameter combinations. A variable-rate fertilising system is presented as an application case for collaborative generation of control index table with the combined accuracy, energy saving and fertilising-consistency optimisation objectives. The experimental results show the effectiveness of the proposed hybrid approach.