Neural State Prediction for Wireless Feedback Control Latency

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing feedback control systems face challenges in migrating to wireless networks for systems with fast dynamics and high reliability requirements, as wireless networks introduce latency that can lead to performance degradation or instability, particularly in Class B and C controlled systems.

Innovation Solution

The implementation of a predictive wireless feedback control system using a state prediction neural network that accounts for communication latencies in wireless networks, enabling the use of wireless time-sensitive networks to synchronize data and predict future states of the target system, allowing for timely control signal generation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If wireless networks are used to convey control signals and measurements, then installation flexibility and mobility are improved, but latency increases leading to performance degradation or instability

Engineering Contradiction:
Improveinstallation flexibilityVSAvoidsystem stability
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent applies preliminary action by predicting the future state of the controlled system before the control signal is applied. The neural network predicts system state at time t+Δt, where Δt is the total latency, so that the control signal is based on anticipated future conditions rather than outdated past measurements, thereby maintaining stability in wireless control systems

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces a neural network as an intermediary component between the measurement receiver and the control signal generator. This neural network processes delayed measurements and predicts future system states, acting as a mediator that compensates for wireless network latency and enables reliable control despite the wireless connection

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If wireless networks are used in control systems, then device mobility is improved, but control precision deteriorates due to latency

Engineering Contradiction:
Improvedevice mobilityVSAvoidcontrol precision
Core Design Contradiction:
Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The system performs preliminary prediction of the controlled system's state at the time the control signal will be effective. By predicting state at t+Δt where Δt accounts for wireless latency, the control precision is maintained despite using mobile wireless connections instead of wired connections

Inventive Principle:
Principle #10Preliminary action

3Reliability

If neural networks are used to predict system states, then control performance in wireless systems is improved, but computational complexity increases

Engineering Contradiction:
Improvecontrol performanceVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent uses a neural network to create a computational model (copy) of the controlled system's dynamics. This neural network model replicates the system's behavior to predict future states, enabling accurate control predictions without requiring complex real-time simulations or iterative calculations during actual control operation

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3757687B1Wireless feedback control loops with neural networks to predict target system states
Publication Date: 2023.11.08 INTEL CORP
  • EP3757687B1 patent drawingFigure 1
  • EP3757687B1 patent drawingFigure 2
  • EP3757687B1 patent drawingFigure 3

AI summary

Example wireless feedback control systems disclosed herein include a receiver to receive a first measurement of a target system via a first wireless link. Disclosed example systems also include a neural network to predict a value of a state of the target system at a future time relative to a prior time associated with the first measurement, the neural network to predict the value of the state of the target system based on the first measurement and a prior sequence of values of a control signal previously generated to control the target system during a time interval between the prior time and the future time, and the neural network to output the predicted value of the state of the target system to a controller. Disclosed example systems further include a transmitter to transmit a new value of the control signal to the target system via a second wireless link.