With the widespread use of nonlinear load in the power grid, the harmonic has seriously affected the quality of power supply and power grid security. In order to effectively control and governance harmonics, we must firstly define the distribution and state of the harmonic in the power grid. The harmonic state estimation in power system is the base of the harmonic control, it can accord the measured value of the limited point to estimate the whole network of harmonic state.Based on analyzing and summarizing to the existing of harmonic state estimate, it is found that there are two big problems in the traditional methods of current researches, the one is not for the optimal placement of meter bus, and the other one is not considering the parameter errors in the harmonic state estimation. This paper studies the above two problem.Firstly, according to the neglect of the optimal placement, this paper gives the optimal placemen for meter bus in the harmonic state estimate based on particle swarm optimization algorithm. Installing different numbers of harmonic measuring device, we can get the optimal placement, which is the research thought in this paper. The objective function for the optimal placemen is the trace of estimation error covariance matrix. In addition, this paper improves the insufficient ofσvalues in objective function. Example varies the correctness of this algorithm.Secondly, aiming at the practical aspects of optimal placement, this paper improves it and puts forward the comprehensive optimal placement, uses the AHP to gain harmonic weight, and establishes the comprehensive objective function. Firstly, this paper selects the trace of estimation error covariance matrix as the objective function for gaining every harmonic optimal placement. Secondly, chooses average trace value change as the objective function for gaining the comprehensive optimal placement for all of kinds of harmonics, and then uses particle swarm optimization algorithm to solve the problem. This paper uses the IEEE-39 and IEEE-145 node system as examples, and varies the rationality of this algorithm.Finally, doing harmonic state estimation based on meter optimal placement, and uses the total least-square to do harmonic state estimation, which takes measurement error and parameters error into account. The IEEE-14 node system is used as example, the results show that this algorithm has good effect. |