Error data analyzing of the integration of the sensors with different measuring boundaries in dynamic environments

Abstract

Data accuracy is important to determine a target. Mobile robots are used in dynamic environments.At dynamic environments, the noise factor and situation of objects are changeable. A sensor may notwork properly in dynamic environmenst. The noise factor affects getting accurate data negatively.These factors depend on sensor types. Therefore, one of the goals of sensor integration i, to quenchone or more sensor disadvantages by using other sensor or sensors, get more accurate and reliabledata, reduce noise factor, determine and analyze more options of target or targets.Every sensor has a measurement range and this range changes according to the sensor types. Thesensor has two boundaries; lower boundary and upper boundary. It is usually zero below the lowerboundary. Above the upper boundary it keeps the maximum output value and Is usually stable.Essentially, the sensor data over the range called error data. If the value to be measured in anintegration composed of the sensors is out of ranges of some sensors, the data of these sensors is to beeliminated. Eliminated data are bad data. Error data analyzing methods are used to determine baddata. Chauvenet’s criterion is one of these methods. A new model eliminating the sensor data isdeveloped. The logic equations are used in Chauvenet’s criterion in this model to determine the datais over range or not. The logic equations are used in Chauvenet’s criterion in this model to determineif the data is over range or not.