Worst-case satisfaction of STL specifications using feedforward neural network controllers: A lagrange multipliers approach

Engineering & Materials Science

• Lagrange multipliers 100%
• Temporal logic 99%
• Feedforward neural networks 93%
• Specifications 54%
• Controllers 45%
• Recurrent neural networks 29%
• Reinforcement learning 28%
• Nonlinear systems 23%
• Neural networks 17%