Future State Estimation Using Infinite-Horizon Probability Distributions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting future states in model predictive control, such as those described in PTLs 1, 2, 3, and 4, require longer calculation times as the prediction horizon increases, limiting their ability to estimate states in an infinite time frame due to computational constraints.
Innovation Solution
A future state estimation device and method that calculates future states of a control object and its environment in an infinite time frame using a probability density distribution, employing a model storage unit, a future state prediction result storage unit, and a future state prediction arithmetic unit to perform calculations equivalent to a series, with an attenuation type state transition matrix that weights distant future transitions, allowing for rapid estimation within a finite state space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iterative calculation method is used to predict future states, then prediction accuracy is improved, but calculation time increases significantly as prediction horizon extends
Solution Approach 1:
The patent pre-calculates and stores transition probability matrices for all possible state transitions in advance. During runtime, future state prediction is achieved by multiplying current state probability distributions with pre-computed transition matrices, avoiding iterative simulation. This preliminary preparation enables rapid prediction at any future time point without increasing calculation time proportionally to prediction horizon.
Solution Approach 2:
The patent transitions from static iterative simulation to dynamic probability distribution multiplication. By representing system states as probability distributions and transitions as stochastic matrices, the method dynamically computes future states through efficient matrix operations rather than step-by-step iteration, significantly reducing calculation time while maintaining accuracy.
2Duration of action of moving object
If prediction horizon is extended to infinite time frame, then long-term control optimization is improved, but computational complexity increases beyond feasible limits
Solution Approach 1:
The patent changes the fundamental parameter representation from deterministic state trajectories to probability distributions. By modeling states stochastically and using probability theory, the system can compute infinite-horizon predictions through convergent matrix series operations rather than infeasible infinite iterative simulations, reducing computational complexity to manageable levels.
Solution Approach 2:
The patent replaces the mechanical iterative simulation process with a mathematical probability-based computation system. Instead of mechanically stepping through time iterations, the system uses probability distribution multiplication and matrix operations to directly compute future states, substituting computational mechanics with efficient mathematical operations.
3Reliability
If conventional model predictive control is used, then control performance is maintained, but calculation speed decreases for long-term predictions
Solution Approach 1:
By pre-computing and storing transition probability matrices for all state transitions, the system eliminates the need for repeated iterative calculations during control execution. This preliminary preparation maintains control performance while enabling rapid prediction at any future time point through simple matrix multiplication operations.
Data Source
AI summary
An objective of the present invention is to provide a future state estimation method and future state estimation device with which, given that the space in which the estimation is carried out is finite, it is possible to rapidly estimate the states of a controlled object and the peripheral environment thereof in the form of probability density distribution for infinite future. Provided is a future state estimation device characterized by comprising: a model storage part for saving a model for simulating a subject of simulation and the peripheral environment of the subject of simulation; a future state forecast result storage part for storing information obtained by estimating future states of the subject of simulation and the peripheral environment of the subject of simulation within a finite space in the form of probability density distribution, for either an infinite time or a given time step in the future; and a future state forecast computation part for carrying out calculation, which is equivalent to a series, using the model for simulating the future states of the subject of simulation and the peripheral environment of the subject of simulation in the form of probability density distribution.