Deep Reinforcement Learning Control of Quantum Cartpoles

Abstract

We generalize a standard benchmark of reinforcement learning, the classical cartpole balancing problem, to the quantum regime by stabilizing a particle in an unstable potential through measurement and feedback. We use state-of-the-art deep reinforcement learning to stabilize a quantum cartpole and find that our deep learning approach performs comparably to or better than other strategies in standard control theory. Our approach also applies to measurement-feedback cooling of quantum oscillators, showing the applicability of deep learning to general continuous-space quantum control.

arXiv comment5+4 pages, 2+2 figures, 2+2 tables, 5 videos at an external link

References in corpus (29)

• Adam: A Method for Stochastic Optimization
• Cavity Optomechanics
• Prioritized Experience Replay
• Dueling Network Architectures for Deep Reinforcement Learning
• The Arcade Learning Environment: An Evaluation Platform for General Agents
• Shortcuts to adiabaticity: concepts, methods, and applications
• Benchmarking Deep Reinforcement Learning for Continuous Control
• Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
• Simple pulses for elimination of leakage in weakly nonlinear qubits
• Time Optimal Control in Spin Systems
• OpenAI Gym
• Quantum control theory and applications: A survey
• Noisy Networks for Exploration
• Feedback-control of quantum systems using continuous state-estimation
• Measurement-based quantum control of mechanical motion
• Optimal control technique for Many Body Quantum Systems dynamics
• Continuous quantum measurement and Itô formalism
• Reinforcement Learning in Different Phases of Quantum Control
• Reinforcement Learning with Neural Networks for Quantum Feedback
• Optimizing Quantum Error Correction Codes with Reinforcement Learning
• Deep Reinforcement Learning for Quantum Gate Control
• Exploring the Quantum Speed Limit with Computer Games
• Evolutionary Algorithms for Hard Quantum Control
• Coherent Transport of Quantum States by Deep Reinforcement Learning
• Global optimization of quantum dynamics with AlphaZero deep exploration
• Reinforcement learning for autonomous preparation of Floquet-engineered states: Inverting the quantum Kapitza oscillator
• Linear quantum trajectories: Applications to continuous projection measurements
• Deterministic generation of remote entanglement with active quantum feedback
• LaProp: Separating Momentum and Adaptivity in Adam

Cited by in corpus (37)

• Artificial Intelligence and Machine Learning for Quantum Technologies
• High-Fidelity, Frequency-Flexible Two-Qubit Fluxonium Gates with a Transmon Coupler
• Model-Free Quantum Control with Reinforcement Learning
• Deep learning enhanced Rydberg multifrequency microwave recognition
• Measurement Based Feedback Quantum Control With Deep Reinforcement Learning for Double-well Non-linear Potential
• Deep Reinforcement Learning for Quantum State Preparation with Weak Nonlinear Measurements
• Quantum enhancements for deep reinforcement learning in large spaces
• Optimal control of quantum thermal machines using machine learning
• Faster State Preparation across Quantum Phase Transition Assisted by Reinforcement Learning
• Deep reinforcement learning for universal quantum state preparation via dynamic pulse control
• Statistically Characterising Robustness and Fidelity of Quantum Controls and Quantum Control Algorithms
• Detection of entangled states supported by reinforcement learning
• Deep reinforcement learning for preparation of thermal and prethermal quantum states
• Universal quantum state preparation via revised greedy algorithm
• Preparing Schrödinger cat states in a microwave cavity using a neural network
• Addressing the Non-perturbative Regime of the Quantum Anharmonic Oscillator by Physics-Informed Neural Networks
• Enhanced quantum state preparation via stochastic prediction of neural network
• Accelerated Magnonic Motional Cooling with Deep Reinforcement Learning
• Entanglement engineering of optomechanical systems by reinforcement learning
• Reinforcement learning assisted non-reciprocal optomechanical gyroscope
• Designing fast quantum gates using optimal control with a reinforcement-learning ansatz
• Reinforcement Learning for Rotation Sensing with Ultracold Atoms in an Optical Lattice
• Quantum optimal control in quantum technologies. Strategic report on current status, visions and goals for research in Europe
• Characterization of a driven two-level quantum system by Supervised Learning
• Convergent and Efficient Deep Q Network Algorithm
• Deep reinforcement learning for feedback control in a collective flashing ratchet
• Controlling nonergodicity in quantum many-body systems by reinforcement learning
• Adaptive Random Quantum Eigensolver
• Quantum feedback control with a transformer neural network architecture
• No-Collapse Accurate Quantum Feedback Control via Conditional State Tomography
• Reinforcement learning to learn quantum states for Heisenberg scaling accuracy
• Fingerprint Recognition of Partial Discharge Signals in Deep Learning Enhanced Rydberg Atomic Sensors
• Topological Learning in Multi-Class Data Sets
• Preparation of Many-body Ground States by Time Evolution with Variational Microscopic Magnetic Fields and Incomplete Interactions
• Invariant-based inverse engineering for balanced displacement of a cart-pole system
• Learning Feedback Mechanisms for Measurement-Based Variational Quantum State Preparation
• The Quantum Cartpole: A benchmark environment for non-linear reinforcement learning