A Theory Research Based On Border Sensing

Nowadays you can see so many sensors everywhere. Sensors are widely used in industry, agriculture and many other fields, they are playing a more and more important role in this society. Especially in the measurement field, the measurements are always finished with the help of sensors. As an important index of sensor s, if the resolution can be improved, it will be very significant in the measurement field.As the sensing resolution of sensors is limited, then for each measured signal there will be a ranging area, what we called sensing fuzzy area. If you don't analysis the fuzzy area, the size of it will make measurement error. No matter how large the fuzzy area is, if the area itself is stable, the measurement precision can be improved significantly by confirming the border position of the steady fuzzy area and the d istance from the measured quantity to the border.When the sensor is made, the resolution of itself is confirmed, it's very difficult to improve the resolution index by using some technology or techniques directly. However, when the sensor is used to measure, the measurement is not completed just by using sensors, but also by using A/D?MC U, etc. So an equivalence theory that considering the measurement system that made up of sensors, A/D, MC U as a sensor can be used. Then what we need to do is how to use the border sensing theory to improve the resolution index of the equivalence sensor, which can improve the final measurement precision. In fact the actual sensors can be divided into digital sensors and analog sensors. When using the equivalence sensor composed of the digital sensor, as the output of the digital sensor is quantitative value, and the border stability of the quantification is higher than the quantitative resolution itself, so the border sensing theory can be used, after a further curving compensation, the final measurement precision can be improved. When using the equivalence sensor composed of the analog sensor, similarly, as the output of the analog sensor is continuous signal, the output should be processed by AD into digital value to show the measurement result. Then the border sensing theory and some curving compensation methods can be also used when dealing with the A/D conversion to improve the measurement precision of the equivalence sensor.Based on the above thoughts, a measurement s ystem is devised in this paper, here the AD22100 analog sensor is chosen to use. After the experiment, it can be found that using A/D to process the output of the analog sensor, the border sensing theory can be used to improve the quantization precision. In the experiment, it's improved 1 to 2 scales, then the measurement resolution precision of the final measurement system is improved.