Parallel Test Task Scheduling Based On Simulated Annealing Tabu Search Genetic Algorithm

The biggest advantage of parallel test technology is the ability to complete multi-test-task at the same time. Parallel test technology increases the number of chips that can be tested through a period of time, so as to improve the efficiency and throughput of the whole test system; it can reduce the idle time of test resources, in order to improve resource utilization rate; it can also share expensive test resources, thus saving the cost of test. The test task scheduling strategy is a key problem of improving the efficiency of the parallel test system, it is a complicated non deterministic complete problem, and intelligent algorithms can be applied to solve this problem. To solve the complex problem of parallel test task scheduling in the parallel automatic test system, a new static test task scheduling algorithm based on Stochastic Genetic Algorithm, Simulated Annealing Algorithm and Tabu Search Algorithm is proposed. The main ideas of Simulated Annealing Algorithm and Tabu Search Algorithm are applied to Stochastic Genetic Algorithm to avoid premature convergence problem of Genetic Algorithm, so that test task sequences with shortest total test time and highest parallel efficiency can be obtained. This new test task scheduling algorithm has low algorithm complexity, and it can get many efficient parallel test sequences in short time. It can also avoid local optimal solutions and gradually convergent to the feasible globally optimal solution. The experimental result shows that the new test task scheduling algorithm is feasible and ascendant.