Behaviour-based control architecture has successfully demonstrated their competence in mobile robot development. One key issue in behaviour-based design is the action selection problems. In behaviour-based system, a composite behaviour is implemented as a system using Context Dependent Blending (CDB) that activates the underlying individual behaviours according to the current robot's context in a certain degree. However, the compromises of conflicting behaviours decision might be sub-optimal or even worse than any of the individual commands. It is caused by using the un-optimized fuzzy context rules. Therefore, most of the works in the field generate a certain interest for the study of fuzzy systems with added learning capabilities for best fuzzy context rules. This paper presents the development of CDB with Flexible Fuzzy Context Rules (FFCR) using Particle Swarm Optimization (PSO) called as Particle Swarm Fuzzy Controller (PSFC). Several experiments with MagellanPro mobile robot have been performed to analyse the performance of the algorithm. A set of Fuzzy Context Rules, called as Single Fuzzy Context Rules (SFCR) are used as comparison. The promising results have proved that the proposed control architecture for mobile robot has better capability to accomplish special task in office-like environment.