Real-time optimization using reinforcement learning

Powell, Kody M.; Machalek, Derek; Quah, Titus

doi:10.1016/j.compchemeng.2020.107077

Cited by 60 publications

(15 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other improvements to ANN-PSO include decreasing the computational time. Our previous work shows replacing PSO with an actor ANN can significantly reduce the online computational time and find an optimal policy [48]. As PPO and other on-policy algorithms requires interaction with an environment, they should be explored in their ability to augment FP-NLP.…”

Section: Discussionmentioning

confidence: 99%

Comparing Reinforcement Learning Methods for Real-Time Optimization of a Chemical Process

2020

Self Cite

View full text Add to dashboard Cite

One popular method for optimizing systems, referred to as ANN-PSO, uses an artificial neural network (ANN) to approximate the system and an optimization method like particle swarm optimization (PSO) to select inputs. However, with reinforcement learning developments, it is important to compare ANN-PSO to newer algorithms, like Proximal Policy Optimization (PPO). To investigate ANN-PSO’s and PPO’s performance and applicability, we compare their methodologies, apply them on steady-state economic optimization of a chemical process, and compare their results to a conventional first principles modeling with nonlinear programming (FP-NLP). Our results show that ANN-PSO and PPO achieve profits nearly as high as FP-NLP, but PPO achieves slightly higher profits compared to ANN-PSO. We also find PPO has the fastest computational times, 10 and 10,000 times faster than FP-NLP and ANN-PSO, respectively. However, PPO requires more training data than ANN-PSO to converge to an optimal policy. This case study suggests PPO has better performance as it achieves higher profits and faster online computational times. ANN-PSO shows better applicability with its capability to train on historical operational data and higher training efficiency.

show abstract

Section: Discussionmentioning

confidence: 99%

Comparing Reinforcement Learning Methods for Real-Time Optimization of a Chemical Process

2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…To reach such performance, a reward function is required, which transforms process targets into rewards, allowing to learn the optimal policy. The reward design can be based on different target variables, such as real-time profits (Powell, Machalek, and Quah 2020), cost-per-time function (Quah, Machalek, and Powell 2020), or similarity measures based on specified performance criteria (He et al 2020). The individual goal-oriented design enables a broad application in further applications such as flotation processes to reduce non-dynamic drawbacks of modelbased approaches (Jiang et al 2018), in laser welding to increase process repeatabilities (Masinelli et al (2020), and others), or in injection molding to broaden up narrow process windows of conventional methods in ultrahigh precision processes .…”

Section: Process Controlmentioning

confidence: 99%

Deep reinforcement learning in production systems: a systematic literature review

Panzer

Bender

2021

International Journal of Production Research

130

View full text Add to dashboard Cite

Shortening product development cycles and fully customisable products pose major challenges for production systems. These not only have to cope with an increased product diversity but also enable high throughputs and provide a high adaptability and robustness to process variations and unforeseen incidents. To overcome these challenges, deep Reinforcement Learning (RL) has been increasingly applied for the optimisation of production systems. Unlike other machine learning methods, deep RL operates on recently collected sensor-data in direct interaction with its environment and enables real-time responses to system changes. Although deep RL is already being deployed in production systems, a systematic review of the results has not yet been established. The main contribution of this paper is to provide researchers and practitioners an overview of applications and to motivate further implementations and research of deep RL supported production systems. Findings reveal that deep RL is applied in a variety of production domains, contributing to data-driven and flexible processes. In most applications, conventional methods were outperformed and implementation efforts or dependence on human experience were reduced. Nevertheless, future research must focus more on transferring the findings to real-world systems to analyse safety aspects and demonstrate reliability under prevailing conditions.

show abstract

“…Offline character identification execution is difficult because of its variation of fonts and writers. The study illustrates a higher accuracy rate for identification of isolated words and characters in printed text/optical character recognition (OCR); but, it is necessary for a competent handwritten character detection scheme able to generate a higher degree of accuracy in handwritten text detection [2].…”

Section: Introductionmentioning

confidence: 99%

An Intelligent Telugu Handwritten Character Recognition Using Multi-Objective Mayfly Optimization with Deep Learning–Based DenseNet Model

Sonthi

Nagarajan

Krishnaraj

2023

ACM Trans. Asian Low-Resour. Lang. Inf. Process.

View full text Add to dashboard Cite

Handwritten character recognition process has gained significant attention among research communities due to the application in assistive technologies for visually impaired people, human robot interaction, automated registry for business document, and so on. Handwritten character recognition of Telugu language is hard owing to the absence of massive dataset and trained convolution neural network (CNN). Therefore, this paper introduces an intelligent Telugu character recognition using multi-objective mayfly optimization with deep learning (MOMFO-DL) model. The proposed MOMFO-DL technique involves DenseNet-169 model as a feature extractor to generate a useful set of feature vectors. Moreover, functional link neural network (FLNN) is used as a classification model to recognize and classify the printer characters. The design of MOMFO technique for the parameter optimization of DenseNet model and FLNN model shows the novelty of the work. The use of MOMFO technique helps to optimally tune the parameters in such a way that the overall performance can be improved. The extensive experimental analysis takes place on benchmark datasets and the outcomes are examined with respect to different measures. The experimental results pointed out the supremacy of the MOMFO technique over the recent state of art methods.

show abstract

Real-time optimization using reinforcement learning

Cited by 60 publications

References 35 publications

Comparing Reinforcement Learning Methods for Real-Time Optimization of a Chemical Process

Comparing Reinforcement Learning Methods for Real-Time Optimization of a Chemical Process

Deep reinforcement learning in production systems: a systematic literature review

An Intelligent Telugu Handwritten Character Recognition Using Multi-Objective Mayfly Optimization with Deep Learning–Based DenseNet Model

Contact Info

Product

Resources

About