Research On Theoretical Line Loss Calculation By State Estimation And Optimization Algorithm

Theoretical line loss rate is an important economic indicator in the operation of power system, and is also an important symbol as a measure of the level of power supply enterprise management. But there may be bad data among the actual measurements, which affects the accuracy of theoretical line loss calculation. State estimation can provide a reliable database for the theoretical line loss calculation, which is an important means that can ensure measurement's accuracy and completeness. So it makes the results of the theoretical line loss calculation realer.According to the needs of theoretical line loss calculation, this paper researches on the methods of removing bad data out of the system and selecting the node type in theoretical line loss calculation by applying optimization algorithms and state estimation. Thereby, these above can enhance the accuracy and ensure the reliability of theoretical line loss calculation.First of all, the problem is researched on that the different combinations of PQ, PV node-type in the theoretical line loss calculation affect the accuracy of result. In the power system's measure system, the bus voltage and injective reactive power can generally be obtained at the same time, but during calculating only one can be chosen. In this paper, state estimation is used to remove bad data, and then the high-precision data are selected by particle swarm optimization (PSO), in the end the optimal combination of PQ and PV node-type is decided. In addition particle swarm optimization algorithm is improved by some appropriate improvements for the calculation.Second, a new method of bad data identification is studied to excluding bad data among the original data. The virtual path is proposed here to combine ant colony optimization (ACO) and sensitivity analysis. This method selects some good data which are part of all, and the use these data to identify the others with the method of adding one-dimensional data. At last the collection of bad data is received. The experiments show the method in this paper has a good effect on bad data identification.Finally, it is studied that the overall approach to select the balance node PQ and PV node-type. The use of ACO and evidence theory is to achieve the balance node-type's selection, and then part of all the measurement which are better than others are selected. So the other measurements go through the processing of bad data detection and identification, after that the method above of PSO is used to decide the node-type of PQ and PV. In the end it is achieved that the precise calculation of theoretical line loss.In this paper, IEEE-14 and IEEE-39 bus system is used as examples to check the effect of the simulations by Java and Matlab, the results obtained show that the proposed algorithm has good results.