基于单神经元的智能温度控制仪的研制

Development of Intelligent Temperature Controller Based on the Single Neuron

温度对象种类繁多 ,普遍具有时间常数大、纯滞后时间长、时变性较明显等特点 ,这使得基于被控对象数学模型的常规控制方法、模糊控制方法及专家智能控制等方法在温度控制中有较大的局限性而无法获得理想的控制效果。针对以上情况 ,本文采用具有自学习能力的单神经元自适应智能控制算法来对温度对象实施控制 ,该算法简单 ,易于实现。最后 ,研制出基于该算法和 16位单片机 (80C196KB)的新型智能温度控制仪 ,经仿真及实际应用证明 。

The temperature object is of a great variety and it usually has the features of large inertia, severe delay and considerable time variation. Those ways like fuzzy control, intelligent control and other conventional control algorithms based on the mathematical model of controlled device can't get the ideal control results because of their limitation. To tackle the above control difficulties, the adaptive intelligent control algorithm of single neuron, which has the self learning ability and is simple to calculate and easy to implement, is applied. Finally a new type of intelligent temperature controller based on the adaptive intelligent control algorithm of single neuron and 16 bit single chip computer (80C196KB) is developed. After simulation and practical test, it was demonstrated that the new temperature controller has the better control effect and application prospects.