The current microfluidic platform has a single function and cannot meet the needs of a variety of related experiments. In this paper, a universal droplet manipulation platform is designed. Taking the theory of electric wetting as the basis of droplet driving, an arrayed reconfigurable microfluidic chip is designed using a printed circuit board to realize the control of droplets. An online detection method based on improved particle swarm algorithm is designed to detect faulty electrodes. Computer vision technology and image processing algorithms are utilized to locate the droplets and obtain droplet information. The platform can realize an average droplet driving speed of 27.27 mm/s. The online detection algorithm successfully finds an optimal path with a length of 822 units in the test, and the accuracy of droplet position detection can reach more than 90%.