Abstract
Optimization problems are ubiquitous in scientific research, engineering, and daily lives. However, solving a complex optimization problem often requires excessive computing resource and time and faces challenges in easily getting trapped into local optima. Here, we propose a memristive optimizer hardware based on a Hopfield network, which introduces transient chaos to simulated annealing in aid of jumping out of the local optima while ensuring convergence. A single memristor crossbar is used to store the weight parameters of a fully connected Hopfield network and adjust the network dynamics in situ. Furthermore, we harness the intrinsic nonlinearity of memristors within the crossbar to implement an efficient and simplified annealing process for the optimization. Solutions of continuous function optimizations on sphere function and Matyas function as well as combinatorial optimization on Max-cut problem are experimentally demonstrated, indicating great potential of the transiently chaotic memristive network in solving optimization problems in general.
Highlights
People are inevitably facing various optimization problems at all times to improve efficiency, use resources rationally, and find the best solution under certain constraints
It is worth mentioning that an extra summation term is needed after the mathematical transformation, the most computationally intensive part is still the VMM in the second term, which can be calculated by the memristor array tions in a single read operation
A network of needs about (2n2 − 1)nt operations from the second term, whereas the rest in Eq 7 only requires about n2 + nt − 1
Summary
People are inevitably facing various optimization problems at all times to improve efficiency, use resources rationally, and find the best solution under certain constraints. Heuristic algorithms inspired by human intelligence, animal society, and physical phenomena, such as artificial neural network, genetic algorithm, ant colony algorithm, and simulated annealing [2], have been developed to tackle these problems. Among these approaches, the Hopfield network [3] can solve optimization problems by minimizing its energy function during network evolution and has been considered suitable for efficient hardware implementation because of its simple computing elements and parallel computing process. Combination of simulated annealing with Hopfield networks could further help the network jump out of local minima and find a better or even optimal solution [5]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.