Evaluation and error minimization of dynamic short time load forecasting model with control charts and process capability analysis in the presence of distributed generation

Short time load forecasting (STLF) is a pivotal concept in energy marketing, therefore, regulatory has defined penalty for load forecasting errors, which disturb energy market balance. Distributed generation (DG) has two effects on STLF models: first, in the presence of DG these models inevitably entail non-repeating data as well as load trends, and second, the share of DG in power generation is not constant. Therefore, the STLF model should be evaluated and improved continuously otherwise model accuracy will dwindle gradually. A lot of STLF models have been developed but there isn´t proper tool to assess their accuracy in the presence of DG. For controlling the impact of probabilistic behaviour of distributed generators on load forecasting, West Tehran province power distribution company (WTPPDC) combined dynamic model, statistical control chart and Process capability analysis for continual evaluation and monitoring of the STLF model. In this study WTPPDC have used process capability analysis for evaluation of forecasting capability of model and control charts for detecting out of control error and accumulative bias in prediction in the presence of DG. Quality approach to load forecasting error controlling can help distribution companies to improve their model before forecasting errors reduce their profit and business confidence.