Extending Taguchi's Approach to Adaptive Stochastic Control: a Simplified Model for Feedback Adjustments of both Mean and Variance