Seven dubious methods to compute optimal Q for Bayesian stochastic particle flow

We describe several new algorithms for particle flow using non-zero diffusion with optimal Q as well as suboptimal Q. The purpose of using optimal Q is two fold: improve estimation accuracy of the state vector and improve the accuracy of uncertainty quantification. Q is the covariance matrix of the diffusion for particle flow corresponding to Bayes' rule.