Development of a smart energy saving device based on Ant Colony System

The purpose of this paper is to introduce an economic cooperative agent based optimization algorithm oriented on Ant Colony System (ACS), which is sensitive to load profile changes, for controlled and optimized switching of Static VAR Compensators (SVCs). Compared to other available heuristics ACS offers several advantages including distributed computation, positive feedback, and robustness. Optimization parameters used for the present method includes power factor, voltage profile, frequency and the health and state of capacitors and also the instant of switching. The smart device, implemented using BF533 DSP microcontroller incorporating this algorithm is to be used for traction substations, rolling mills, cold storages and furnaces with dynamic load pattern. This device will reduce the penalty charged to the consumers and also decrease loading of cables and other electrical apparatus connected to the system. Comparison of results show that ACS based algorithm, in contrast to Genetic Algorithm (GA), can offer optimum solution in much lesser time with less memory requirement. The DSP hardware platform, implementing ACS based method took 102 milli sec for searching a combination. The same system using GA based method took 3 min. 34 sec. for converging to the same result.