Development of an operation trajectory design algorithm for control of multiple 0D parameters using deep reinforcement learning in KSTAR

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

This work develops an artificially intelligent (AI) tokamak operation design algorithm that provides an adequate operation trajectory to control multiple plasma parameters simultaneously into different targets. An AI is trained with the reinforcement learning technique in the data-driven tokamak simulator, searching for the best action policy to get a higher reward. By setting the reward function to increase as the achieved β _p, q ₉₅, and l _i are close to the given target values, the AI tries to properly determine the plasma current and boundary shape to reach the given targets. After training the AI with various targets and conditions in the simulation environment, we demonstrated that we could successfully achieve the target plasma states with the AI-designed operation trajectory in a real KSTAR experiment. The developed algorithm would replace the human task of searching for an operation setting for given objectives, provide clues for developing advanced operation scenarios, and serve as a basis for the autonomous operation of a fusion reactor.

Related collections

Most cited references 17

Record: found
Abstract: found
Article: not found

Long Short-Term Memory

Jürgen Schmidhuber, Jürgen Schmidhuber (2002)

Learning to store information over extended time intervals by recurrent backpropagation takes a very long time, mostly because of insufficient, decaying error backflow. We briefly review Hochreiter's (1991) analysis of this problem, then address it by introducing a novel, efficient, gradient-based method called long short-term memory (LSTM). Truncating the gradient where this does not do harm, LSTM can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units. Multiplicative gate units learn to open and close access to the constant error flow. LSTM is local in space and time; its computational complexity per time step and weight is O(1). Our experiments with artificial data involve local, distributed, real-valued, and noisy pattern representations. In comparisons with real-time recurrent learning, back propagation through time, recurrent cascade correlation, Elman nets, and neural sequence chunking, LSTM leads to many more successful runs, and learns much faster. LSTM also solves complex, artificial long-time-lag tasks that have never been solved by previous recurrent network algorithms.

0 comments Cited 6691 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: not found

Human-level control through deep reinforcement learning.

Volodymyr Mnih, Koray Kavukcuoglu, David Silver … (2015)

The theory of reinforcement learning provides a normative account, deeply rooted in psychological and neuroscientific perspectives on animal behaviour, of how agents may optimize their control of an environment. To use reinforcement learning successfully in situations approaching real-world complexity, however, agents are confronted with a difficult task: they must derive efficient representations of the environment from high-dimensional sensory inputs, and use these to generalize past experience to new situations. Remarkably, humans and other animals seem to solve this problem through a harmonious combination of reinforcement learning and hierarchical sensory processing systems, the former evidenced by a wealth of neural data revealing notable parallels between the phasic signals emitted by dopaminergic neurons and temporal difference reinforcement learning algorithms. While reinforcement learning agents have achieved some successes in a variety of domains, their applicability has previously been limited to domains in which useful features can be handcrafted, or to domains with fully observed, low-dimensional state spaces. Here we use recent advances in training deep neural networks to develop a novel artificial agent, termed a deep Q-network, that can learn successful policies directly from high-dimensional sensory inputs using end-to-end reinforcement learning. We tested this agent on the challenging domain of classic Atari 2600 games. We demonstrate that the deep Q-network agent, receiving only the pixels and the game score as inputs, was able to surpass the performance of all previous algorithms and achieve a level comparable to that of a professional human games tester across a set of 49 games, using the same algorithm, network architecture and hyperparameters. This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.