Fuzzy control is suitable for automotive control, because of its controllability as good as that obtained by mankind. However, since automotive control requires milli-second response and learning control, and the fuzzy system in automobiles requires small-sized compact ( built-in type ), a custom fuzzy inference LSI is needed for automotive control. We then investigated requirements of a fuzzy inference LSI suitable for automotive control and fabricated a fuzzy inference LSI satisfactory to these requirements.

The fabricated fuzzy LSI is designed to be used in various automotive control experiments. The inference time is 63 micro-seconds under the conditions of six inputs, two outputs and 256 rules. Combination of this fuzzy LSI only with EPROM makes it possible to implement a fuzzy control system. The fuzzy LSI was applied to a rough road durability test, aiming at the automatic driving equivalent to human driver's operation. In the test, the accumulated steering compensation of less than twenty degrees in the fuzzy control was about one third of that in the linear control.

The fuzzy LSI fabricated for use in various experiments is too large ( 10.7 mm × 10.9 mm) to adopt as an automotive part. Therefore, we studied a smaller-sized fuzzy LSI. The obtained chip is small enough ( 4.8 mm × 4.8 mm ) for automotive applications and can be built in a microprocessor as a fuzzy inference co-processor without any other circuits.

�@