Biologically-Inspired Personalities for Control Systems and Robots Using Nonlinear Optimization and Feedback Theory

In this paper we present a method for mathematical and computer modeling of human personalities, in a way that makes it suitable for biologically-inspired control system design. The clinical personality disorders, Lowry Colors, and Enneagram map to a quantitative, graphical quadrant model of human personality self-images. The four qualities on the quadrant system have similarities to the four terms in a cost-functional for control-system design, the four parameters in a PID-plus-bias control system, and the four letters J/P and T/F in the Myers-Briggs behavioural technique personality system. Thus, predicting the Myers-Briggs probability distribution in the human population becomes a matter of performing near-optimization, with many random initial conditions, on the self-image cost-functional. Then, a proposed biological model of personalities shows how one can predict the probability distribution of Myer-Briggs letters N/S and I/E in the human population. Being quantitative and replicable, the model clearly shows the way forward for programming personalities on robots and AI systems using well-established concepts from feedback theory.