Power Quality Prediction at Consumers Using a Hybrid Knowledge-Based Approach

Prediction is the basis of good planning and management in power systems of smart cities, especially when considering sustainability. The same applies for power quality, that in the last ten years has become an issue for both electricity consumers and power systems' operators. This is due to the negative impact of low power quality on sensitive equipment. In the case of consumers, frequently their own equipment is the source of disturbances that decrease the power quality level. This paper presents a hybrid knowledge-based system for power quality issues prediction. The main component of the proposed system is the knowledge base that contains all information about the sources of power quality problems, like nonlinear, and inductive loads. Production rules, classes, and objects represent the knowledge in the knowledgebase. A mathematical block makes the calculus when diverse sources of power quality disturbances overlap or add. The testing of the prediction system by considering diverse scenarios showed reliable results. The proposed system is beneficial for residential consumers, but also prosumers from smart cities.