Selection of Number and Locations of Temperature and Luminosity Sensors in Intelligent Greenhouse

摘要

Using a wireless sensor network (WSN) for the monitoring and control of parameters, like temperature and luminosity, inside an intelligent greenhouse is generally a better option rather than using a single sensor for each parameter. Distribution of the wireless sensors at random locations will generally yield an incorrect measurement of temperature/luminosity profile inside the greenhouse. In addition, it will unnecessarily increase the number of sensors and the complexity of WSN. In this paper, two approaches namely equal sensor-spacing method and trial-and-error method, for the selection of number and locations of wireless sensors nodes, containing temperature and luminosity sensors, in a greenhouse have been successfully investigated. In the equal sensor-spacing method, the individual errors in temperature and luminosity profile measurement as well as the combined error or rss error (root of sum of squares of individual errors) reduce as the number of sensor nodes is increased from 3 to 10. Although the error corresponding to ten sensors is quite small, such a large number of sensor nodes would make the WSN-based measurement system very expensive. In case of trial and error method, the least value of rss error comes out to be 2.2545% for 3 sensors, 1.2720% for 4 sensors and 3.6157% for 5 sensors. An interesting observation for this approach is that increasing the number of sensors does not necessarily reduce the error, and, therefore, an optimal number and optimal locations of sensors need to be determined by trial and error. Comparison of results for the two methods shows that the same order of error can be achieved with trial-and-error method using lesser number of sensors. The proposed approaches can be used for the measurement of profile along both the dimensions, i.e. the length and breadth, of the greenhouse.