Self-powered acceleration sensors arrayed by swarm intelligence for table tennis umpiring system

Table tennis competition is voted as one of the most popular competitive sports. The referee umpires the competition mainly based on visual observation and experience, which may make misjudgments on competition results due to the referee's subjective uncertainty or imprecision. In this work, a novel intelligent umpiring system based on arrayed self-powered acceleration sensor nodes was presented to enhance the competition accuracy. A sensor node array model was established to detect ball collision point on the table tennis table. This model clearly illuminated the working mechanism of the proposed umpiring system. And an improved particle swarm optimization (level-based competitive swarm optimization) was applied to optimize the arrayed sensor nodes distribution by redefining the representations and update rules of position and velocity. The optimized results showed that the number of sensors decreased from 58 to 51. Also, the reliability of the optimized nodes distribution of the table tennis umpiring system has been verified theoretically. The results revealed that our system achieved a precise detection of the ball collision point with uniform error distances below 3.5 mm. Besides, this research offered an in-depth study on intelligent umpiring system based on arrayed self-powered sensor nodes, which will improve the accuracy of the umpiring of table tennis competition.