Test Path Optimization Of Digital Microfluidic Biochip Based On Intelligent Algorithm

Digital microfluidic biochip(DMFB)can realize the function of large laboratory on the miniaturized chip because of its high integration and convenient operation,which makes it gradually show broad application prospects in the fields of biology,medicine and so on.Because there are potential faults in the production and use of the chip,and the application field of the chip has very high requirements for safety,it is necessary to fully test the chip to ensure the reliability of DMFB.During the test,it is not only necessary to cover all possible faults,but also to minimize the test time and improve the test efficiency.In this paper,the catastrophic fault test of the chip is realized by driving the test droplet to traverse all the array units and their boundaries in the test model.Therefore,the testing of DMFB is essentially a test path planning problem.And designing a simple and effective test path optimization method is of great research significance for the application of DMFB.Aiming at the issue that the randomness of test path search and optimization using single intelligent algorithm is large,which leads to the problem that the test efficiency needs to be improved,a hybrid algorithm combining priority strategy and particle swarm optimization algorithm(PSO)is proposed,which shortens the length of test path and improves the test efficiency.The priority strategy is used to generate the initial test path,and the PSO algorithm is used to optimize the priority coefficient in the test model to obtain a shorter test path.This paper realizes the exchange of priority coefficients by changing the particle velocity in PSO algorithm.At the same time,the Kaiser window function is introduced into the inertia weight to replace the conventional linear method in a nonlinear way to improve the search performance of the algorithm.Through the simulation test of 15×15 chip,the results show that the online test efficiency of the hybrid algorithm proposed in this paper is 1.7% higher than that of ant colony algorithm.Moreover,the problem of "deadlock" of droplets in the generation of test path is solved by using backoff operation.In order to further improve the test efficiency,a hybrid algorithm combining priority strategy and genetic algorithm(GA)is proposed in this paper.The hybrid algorithm first obtains the test path of the chip through priority strategy,then uses the GA to realize the selection,crossover and mutation of the priority coefficient,and gradually realizes the test path optimization in an iterative way.Through the simulation test of several chips with different sizes,the results show that the hybrid algorithm improves the online test efficiency by 3.7% compared with PSO algorithm.The proposed hybrid algorithm can obtain the optimal value of the test path whether offline test or online test,which shows that the hybrid algorithm combining priority strategy and GA is an effective method to solve the problem of test path optimization of DMFB.Based on the analysis of the characteristics of universal pin-constrained digital microfluidic biochip(PDMFB),the pin constraint rule of PDMFB with "connect-5" structure is deduced.On the basis of considering the design and online test requirements of PDMFB with "connect-5" structure,this paper puts forward a new pin allocation scheme of PDMFB with "connect-5" structure for the online test.It solves the problem that the chip cannot be tested online under the original pin allocation scheme,so as to improve the reliability of biochemical experiment.At the same time,the hybrid method of priority strategy and GA is used to solve the issue of online test path optimization of PDMFB with "connect-5" structure for the first time,and the length of test path is shortened.Aiming at the problem of parallel test path optimization of DMFB with multiple droplets,a test strategy of serial-parallel conversion is proposed in this paper,which reduces the total amount of test tasks of partition strategy.At the same time,a method of adaptive adjustment of test tasks is proposed to improve the defect brought by the method of average allocation of test tasks.The hybrid algorithm combining priority strategy and genetic algorithm is used to realize the parallel test of chip.The test simulation of different chips shows that the serial-parallel conversion strategy can improve the test efficiency by more than 5.4% compared with the partition strategy.Moreover,compared with the way of average task allocation,the way of adaptive task adjustment further optimizes the online test efficiency of chips.