Research On On-line Testing Method Of Digital Microfluidic Biochip

With the advantage of high flexibility,Digital microfluidic biochip is becoming a promised platform for Lab-on-a-chip and become widespread in safety-critical biochemistry and biomedical applications.However,due to the complexity of manufacturing procedure,Digital Microfluidic Biochip is more susceptible to defects during manufacturing and field operation.Therefore,effective testing methods are required to ensure system dependability.As on-line testing is an essential part of testing techniques,this thesis mainly focus on finding an effective on-line testing method.On-line testing can be performed by routing the test droplet through the chipunder-test to detect potential faulty positions on the chip.Specially,on-line testing runs simultaneously with bioassays on a chip.Therefore,the key of an on-line testing method is to design a practical and efficient test path planning method.To obtain effective on-line testing method,test droplet needs not only visiting every electrodes in the array,but also traversing every two adjacent electrodes.Meanwhile,during the routing of test droplet,normal assays on the chip can not be interpreted.As a result,routing test droplet for on-line testing is actually a routing problem under constraint.By considering fault coverage rate and fluidic constraints limitation,we first define an on-line testing model and propose a testing method based on improved A* algorithm to optimize on-line testing time.We formulate test planning problem in terms of Chinese Postman Problem from graph theory,utilizing real-time search mechanism of A* algorithm to handle the fluidic constraint problem.By improving the evaluation function to follow path construction rules designed for obtaining efficient on-line testing method.To obtain shorter on-line testing time,we further consider concurrent on-line testing with the application of multiple test droplets.Based on the improved A* algorithm,we designed a multiple test droplets on-line testing method.On-line testing time is shorten and the same fault coverage with single droplet on-line testing method is obtained.As some electrodes on the chip are occupied by bioassays for quite long time,we introduce guidance mechanism into improved A* algorithm to allow these positons are earlier detected before they are occupied to optimize on-line testing time.The simulation shows that with single test dropl et,the proposed improved A* algorithm can effectively reduce on-line testing time and enable higher testing efficiency.By using multiple droplets and performing concurrent on-line testing on one chip,testing time is further shortened.The simulation shows that when the occupation time of each partitions slightly varies from each other,we can more effectively reduce on-line testing time.Moreover,when test droplets are guided to detect some long-occupied positions before they are occupied,on-line testing time can be further slightly shortened.However,more test droplets mean more consumption of system resource,the balance between test efficiency and cost should be considered in practical.