An Optimization Method of Sensor Layout to Improve Source Identification Accuracy in the Indoor Environment

This study presents an optimization method of sensor layout to improve identification accuracy of indoor contaminant sources. The method integrates an index, the performance of sensor layout (PSL), with a two-step screening procedure to determine sensor layouts that have potential to achieve relatively high levels of accuracy in source identification. Using the PSL, the performance of each possible sensor layout can be predicted and evaluated, and therefore the optimization method can be performed without running a source identification model. The relationship between source identification accuracy and sensor layout was revealed through case studies in a three-dimensional office. The optimization method was demonstrated and validated by two cases in the same office, which includes optimization of sensor layouts with one and five sensors. The case studies indicate that the presented method can significantly improve source identification accuracy. The influencing factors of optimization results were discussed, and the methods to exclude unexpected optimization results were proposed.