Distributed Derivative-Free Learning Method for Stochastic Optimization Over a Network With Sparse Activity

Li, Wenjie; Assaad, Mohamad; Zheng, Shiqi

doi:10.1109/tac.2021.3077516

Cited by 2 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [9], models were developed to optimize the transportation network of natural gas on the Norwegian Continental Shelf. Oilfield management in long-term operations has also been the focus of derivative-free optimization (DFO) [10], [11], [12], [13]. In [14], for instance, different DFO methods were discussed, including, but not limited to, Hooke-Jeeves direct search, genetic algorithms and particle swarm optimization.…”

Section: A Overview and Related Workmentioning

confidence: 99%

Derivative-Free Optimization With Proxy Models for Oil Production Platforms Sharing a Subsea Gas Network

et al. 2023

View full text Add to dashboard Cite

The deployment of offshore platforms for the extraction of oil and gas from subsea reservoirs presents unique challenges, particularly when multiple platforms are connected by a subsea gas network. In the Santos basin, the aim is to maximize oil production while maintaining safe and sustainable levels of CO 2 content and pressure in the gas stream. To address these challenges, a novel methodology has been proposed that uses boundary conditions to coordinate the use of shared resources among the platforms. This approach decouples the optimization of oil production in platforms from the coordination of shared resources, allowing for more efficient and effective operation of the offshore oilfield. In addition to this methodology, a fast and accurate proxy model has been developed for gas pipeline networks. This model allows for efficient optimization of the gas flow through the network, taking into account the physical and operational constraints of the system. In experiments, the use of the proposed proxy model in tandem with derivative-free optimization algorithms resulted in an average error of less than 5% in pressure calculations, and a processing time that was over up to 1000 times faster than the phenomenological simulator. These results demonstrate the effectiveness and efficiency of the proposed methodology in optimizing oil production in offshore platforms connected by a subsea gas network, while maintaining safe and sustainable levels of CO 2 content and pressure in the gas stream.INDEX TERMS Oil production systems, subsea gas networks, proxy modeling, derivative-free optimization.

show abstract

Section: A Overview and Related Workmentioning

confidence: 99%

Derivative-Free Optimization With Proxy Models for Oil Production Platforms Sharing a Subsea Gas Network

et al. 2023

View full text Add to dashboard Cite

show abstract

“…In our previous work [11], [10], [12], we considered a different framework where each node controls its own action to perform ZOSCO. The network model was also different: each node has a constant probability to communicate with any other node of the network.…”

Section: Introductionmentioning

confidence: 99%

Distributed Zeroth-Order Stochastic Optimization in Time-varying Networks

Li,

Assaad

2021

Preprint

Self Cite

View full text Add to dashboard Cite

We consider a distributed convex optimization problem in a network which is time-varying and not always strongly connected. The local cost function of each node is affected by some stochastic process. All nodes of the network collaborate to minimize the average of their local cost functions. The major challenge of our work is that the gradient of cost functions is supposed to be unavailable and has to be estimated only based on the numerical observation of cost functions. Such problem is known as zeroth-order stochastic convex optimization (ZOSCO). In this paper we take a first step towards the distributed optimization problem with a ZOSCO setting. The proposed algorithm contains two basic steps at each iteration: i) each unit updates a local variable according to a random perturbation based single point gradient estimator of its own local cost function; ii) each unit exchange its local variable with its direct neighbors and then perform a weighted average. In the situation where the cost function is smooth and strongly convex, our attainable optimization error is O(T −1/2 ) after T iterations. This result is interesting as O(T −1/2 ) is the optimal convergence rate in the ZOSCO problem. We have also investigate the optimization error with the general Lipschitz convex function, the result is O(T −1/4 ).

show abstract

Distributed Derivative-Free Learning Method for Stochastic Optimization Over a Network With Sparse Activity

Cited by 2 publications

References 27 publications

Derivative-Free Optimization With Proxy Models for Oil Production Platforms Sharing a Subsea Gas Network

Derivative-Free Optimization With Proxy Models for Oil Production Platforms Sharing a Subsea Gas Network

Distributed Zeroth-Order Stochastic Optimization in Time-varying Networks

Contact Info

Product

Resources

About