Pilots' control behavior exhibited near the controllability limit sometimes appears to be of bang-bang type or of on-off type. In this report is investigated the controllability limit of human pilots when using a on-off controller and is compared with that when using a linear controller having saturation. A first-order unstable system is employed as the controlled element, and step inputs are used. Results show that the controllability limit is determined by the stability of the limit cycle of the on-off control system and that the stability is a function of the pilot's reaction time lag and the magnitude of inputs. It is also shown that for the first-order unstable system linear control gives a wider controllability limit than on-off control does. Investigation of the controllability limit for higher order systems is suggested.

View full abstract

The experimental results of a study of multivariable manual control systems with relay-control switches when a human operator is placed in compensatory tracking tasks are presented. The relay-control switches are compared with (1) linear control sticks, (2) linear control sticks with dead zone, and (3) relay-control switches generating pulses of constant width. From the experiments, following results are obtained:
(i) By using relay-control switches, a human operator not only counts the number of pulses, but also adjusts the pulse-width to get better control performances.
(ii) For a certain high-order controlled element requiring lead compensation, the control performance is improved by using the controllers (controls) by which a human operator easily keeps the manipulated variables zero.

View full abstract

The experiments with readability for numeral figures on road sign. Its readability was determined by recognition distance for ten numerals 0 through 9 and proportion of misreading between numeral figures. Subjects reported it, whether he identifies letters or not, by recording number at each level of distance between Ss and road sign. Results of the present experiments are as follows; 1. Recognition distance increases with the increase of size of the backboard. 2. Proportion of misreading decreases with the increase of the size of backboad. 3. The effect of letter-type of numeral figures upon recognition distance is not evident, but its effect plays an important role for misreading, Especially, proportion of misreading is small in C-type figures.

View full abstract