Optimal meter placement algorithm for state estimation in power distribution networks

- Meter placement for several load levels has been carried out.
- Correlation in power measurement was included in the proposed model.
- Analytical approach for evaluating the state estimator accuracy.
- Computational cost reduction due to the application of analytical approach.

This paper proposes an algorithm to allocate meters in distribution networks to increase the accuracy of the state estimation. This algorithm considers three new issues in the meter placement problem: correlated measurements, multiple load levels and accuracy evaluation of state estimation without using Monte Carlo Simulation (MCS), that is, based on analytical approach. The objective function used in the meters placement problem is associated with the maximization of the accuracy of the state estimator. This objective function is optimized considering constraints for the maximum number of meters that can be installed and for the accuracy of the state estimator. The optimization problem is solved with the binary particle swarm algorithm. The tests results indicated that the meter placement obtained with the proposed algorithm is very accurate for all levels of the load curve. For example, it has been shown that the accuracy of state estimation is four times better when the correlation in power measurements is included in the model. In addition, the computational cost of estimating the accuracy of the state estimate based on the proposed analytical approach is about 47 times lower than that associated with MCS.