Particle-Filter Motion Model Control Under Dynamic Uncertainty

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Model-based control systems face challenges when dealing with uncertain motion models, particularly in scenarios where system dynamics are partially unknown or changing, leading to suboptimal performance or instability.

Innovation Solution

The implementation of a particle filter that includes a motion model with uncertainty, modeled as a Gaussian process, allowing each particle to have its own motion model, weight, and uncertainty measure, enabling the estimation of both the system state and motion model, even when parameters and dynamics are unknown or changing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If adaptive or learning-based MPC is used to estimate unknown parameters, then control performance is improved, but device complexity increases

Engineering Contradiction:
Improvecontrol performanceVSAvoidcontrol system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary estimation module that acts as a mediator between the control system and the uncertain motion model. This module estimates unknown parameters and dynamics using available measurements, providing corrected motion model predictions to the MPC controller without requiring the controller itself to be overly complex. The intermediary handles the complexity of parameter estimation separately, allowing the MPC to focus on optimization.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The control system is segmented into distinct functional modules: a motion model with uncertain parameters, an estimation module that handles parameter identification, and an MPC controller that performs optimization. This segmentation allows each module to specialize in its function, improving overall performance while managing complexity through modular architecture.

Inventive Principle:
Principle #1Segmentation

2Device complexity

If conventional MPC is used with uncertain motion models, then device complexity is kept low, but control performance deteriorates

Engineering Contradiction:
Improvecontrol system complexityVSAvoidcontrol performance
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The system performs preliminary estimation of unknown parameters and dynamics before the MPC optimization step. By pre-processing the uncertain parameters through estimation using available measurements and previous states, the system prepares corrected motion model predictions that the MPC can then use effectively, ensuring good control performance without excessive complexity.

Inventive Principle:
Principle #10Preliminary action

3Adaptability or versatility

If the motion model is updated in real-time to account for changing dynamics, then adaptability is improved, but measurement precision requirements increase

Engineering Contradiction:
Improvemotion model adaptabilityVSAvoidmeasurement accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system implements feedback mechanisms where the estimation module continuously uses available measurements and previous state information to update parameter estimates. This feedback loop allows the motion model to adapt to changing dynamics while being robust to measurement noise, as the estimation process integrates information over time rather than relying on single precise measurements.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3938852B1Model-based control with uncertain motion model
Publication Date: 2024.01.24 MITSUBISHI ELECTRIC CORP
  • EP3938852B1 patent drawingFigure 1A
  • EP3938852B1 patent drawingFigure 1B
  • EP3938852B1 patent drawingFigure 1C

AI summary

A system is controlled using particle filter executed to estimate weights of a set of particles based on fitting of the particles into a measurement model, wherein a particle includes a motion model of the system having an uncertainty modeled as a Gaussian process over possible motion models of the system and a state of the system determined with the uncertainty of the motion model of the particle, wherein a distribution of the Gaussian process of the motion model of one particle is different from a distribution of the Gaussian process of the motion model of another particle. Each execution of the particle filter updates the state of the particle according to a control input to the system and the motion model of the particle with the uncertainty and determines particle weights by fitting the state of the particle in the measurement model subject to measurement noise.