Optimization Of The Number Of Droplets For On-line Measurement Of Digital Microfluidic Chip

Digital microfluidic technology,as an emerging cross technology,has the advantages of high flexibility,simple experimental operation,and saving of chemical reagents.It is currently widely used in multiple biomedical research fields such as parallel immunoassay,drug discovery,and disease diagnosis.These research fields require high reliability of the experimental results of the chip,and sufficient testing is needed to ensure the reliability of the chip.As an important part of ensuring chip yield,online testing has become a research hotspot of digital microfluidic chips.Based on the research of online testing of chips,this paper focuses on the optimization of the number of droplets for online testing of chips.In this paper,based on the study of the fault generation mechanism,fault detection principle and optimization goals of the DMFBs,the online test path model is established using the SPFA algorithm,and a single drop online test scheme based on the improved hybrid leapfrog algorithm is proposed.The test path characteristics of the chip,the taboo judgment strategy and fitness function of the test droplet are designed,and the normal traversal strategy and the avoidance strategy of the test droplet are developed to achieve a single droplet without interfering with the operation of the experimental droplet.The purpose of online test methods for optimization.In order to further reduce the time for testing droplets to test the chip online,and for the low efficiency of single droplet testing,this paper combines the concept of cooperative learning of multi-agents with reinforcement learning on the basis of improved hybrid leapfrog algorithm,and proposes a hybrid leapfrog based reinforcement.The DMFBs multi-drop parallel online test strategy and the online test drop number optimization scheme based on the learning algorithm further optimize the online test time and test drop number.The multi-element detection chip of human body fluid is selected as the simulation verification platform of this paper.The experimental results show that the single drop test scheme based on the improved mixed leapfrog algorithm can not only meet the test requirements of the chip,but also effectively reduce the test time and increase the single drop test time.Single drop test efficiency.When the hybrid leapfrog-type reinforcement learning algorithm is used to perform multi-droplet parallel online testing and optimization of the number of test droplets on the chip,the online test time can be further reduced,and the efficiency of test droplet testing and the number of test droplets can be optimized.