---
title: 'RBPF: Rao-Blackwellised Particle Filtering'
url: https://www.emergentmind.com/topics/rao-blackwellised-particle-filtering-rbpf
type: topic
---

# RBPF: Rao-Blackwellised Particle Filtering

Rao-Blackwellised Particle Filtering (RBPF) is a sequential Bayesian estimation methodology that exploits conditional structure in state-space models to jointly combine sampling-based (Monte Carlo) methods and analytic marginalization, thereby reducing computational variance and improving statistical efficiency compared to standard particle filters. RBPF is particularly suitable when the latent state can be decomposed into “difficult” (typically nonlinear/non-Gaussian) components, which are sampled via particles, and “easy” (conditioned linear-Gaussian or finite-state Markov) components, which are integrated out analytically using optimal filters such as the Kalman filter or hidden Markov model (HMM) filter. In this framework, each particle carries not only a weighted sample, but also an analytic conditional posterior over the marginalized subspace, yielding lower-variance estimators for posterior expectations and facilitating computational tractability in high-dimensional or hybrid systems.

## 1. Structural Principle and State Factorization

At the core of RBPF is the Rao–Blackwell theorem, which asserts that conditioning on sufficient statistics and marginalizing analytically reduces estimator variance. In sequential inference for state-space models, this is operationalized by partitioning the latent state at each time $t$ into two components:
- $x_t^n$: a low-dimensional nonlinear (often non-Gaussian) vector, tracked by Monte Carlo particles;
- $x_t^l$: a linear-Gaussian (or otherwise tractable) vector, for which an optimal recursive filter is available.

The joint filtering distribution is then factorized as
\[
p(x_{0:t}^n, x_t^l | y_{0:t}) = p(x_{0:t}^n | y_{0:t}) ~ p(x_t^l | x_{0:t}^n, y_{0:t})
\]
as detailed in robust position estimation [2506.03537], with $x_t^n$ as position and $x_t^l$ as velocity, and in RBPFs for jump Markov nonlinear systems [1312.0781], Markov-modulated nonlinear dynamic systems [1311.6486], and generic dynamic Bayesian networks [1301.3853].

Conditionally, given a particle path $x_{0:t}^n$, all "easy" latent variables $x_t^l$ can be integrated exactly using e.g., a Kalman filter, HMM filter, or analytic update for model parameters. This results in a filtering approximation where only the “tough” nonlinear components are sampled, while the high-dimensional tractable components are handled analytically, leading to reduced sample degeneracy and lower Monte Carlo variance.

## 2. RBPF Algorithms: Recursion and Weighting

The typical RBPF iteration consists of the following steps (see [2506.03537][1301.3853][1312.0781]):
1. **Prediction**:
   - For each particle $i$, propagate $x_{t-1,i}^n$ according to the model; for dependent components (e.g., position), use the most current analytic estimate (e.g., velocity from the Kalman filter).
   - For each particle, perform an analytic time update for $x_t^l$.

2. **Weight Update**:
   - Compute the likelihood $p(y_t | x_{t,i}^n)$, often integrating analytically over $x_t^l$, leading to
     \[
     w_{t,i} \propto w_{t-1,i} \, p(y_t | x_{t,i}^n)
     \]
   - When observation models are nonlinear, as in GNSS phase-based positioning [2506.03537], the likelihood may involve specialized ambiguity functions or pseudorange-residual-based metrics.

3. **Resampling**:
   - Normalize weights and resample if the effective sample size (ESS) falls below a threshold.

4. **Analytic Correction**:
   - For each particle, update the optimal filter for $x_t^l$ using the latest observation and the current particle’s $x_t^n$.

This structure generalizes to hybrid discrete/continuous models. In Markov-switching systems [1312.0781][1311.6486], a bank of HMMs marginalizes the discrete mode, while each particle carries an associated continuous state trajectory, leading to reduced degeneracy in the discrete state.

## 3. Model Classes and Applications

RBPF is broadly applicable to hybrid and partially tractable state-space models:
- **Mixed linear/nonlinear systems**: e.g., position-velocity models in GNSS positioning, tracking applications, or signal processing, where velocity is linear-Gaussian (filtered with KF) and position is nonlinear [2506.03537][2406.10389].
- **Jump Markov and Markov-modulated systems**: Models with discrete switching regimes (modes), in which regime trajectories are marginalized via an HMM filter and the continuous state is sampled [1312.0781][1311.6486][1707.01311].
- **SLAM and POMDPs**: State-of-the-art localization and mapping, where a robot’s trajectory is sampled while map landmarks or features are marginalized [1301.3853][2409.16392].
- **Parameter-state models**: Joint estimation for state and unknown parameters, either static (marginalized via analytic updates or grid methods [2311.00836]) or stochastic (hierarchically marginalized with RBPF—see noise-adaptive LDS [1504.05723]).
- **Pose tracking and multi-modal estimation**: PoseRBPF marginalizes rotation (discretized SO(3)), tracks translation by particles for real-time 6D detection [1905.09304].
- **Filtering under constraints**: Multi-body manipulation tasks with contact constraints incorporated via QP into the analytic update per particle [2310.04637].

A summary table of representative applications, showing model structure and marginalized subspace, appears below:

| Application Domain            | Sampled State                  | Marginalized State       |
|-------------------------------|-------------------------------|-------------------------|
| GNSS robust positioning [2506.03537]   | 3D position                  | 3D velocity (KF)        |
| Connected vehicle networks [1702.05792]| GNSS biases                  | Vehicle EKFs            |
| Jump Markov/Hybrid SSM [1312.0781]     | Nonlinear continuous state   | Discrete mode (HMM)     |
| SLAM, POMDP planning [2409.16392]      | Robot trajectory, intent     | Map landmarks (KF), continuous substate |
| Extended object tracking [2406.10389]  | Nonlinear pose/shape         | Shape/extent (KF)       |
| Parameter-state SDE [2311.00836]       | State trajectory             | Model parameters        |
| Object pose estimation [1905.09304]    | 3D translation               | 3D rotation (discrete)  |

## 4. Empirical Performance and Theoretical Properties

RBPF provides quantifiable performance gains due to variance reduction in the marginalization of tractable subspaces [2506.03537][1301.3853][2409.16392][1312.0781]. Notable empirical and theoretical properties include:
- **Sample efficiency**: Marginalizing linear-Gaussian or discrete subspaces allows reduced particle counts for equivalent estimation error (e.g., for GNSS positioning, velocity error ≤ 0.1 m/s improves from 79.6% to 83.3%; position error ≤ 0.3 m improves from 54.4% for PF to 68.5% for RBPF [2506.03537]).
- **Variance reduction**: Rao–Blackwellization assures, via central limit theorems (CLT), that the asymptotic variance of RBPF estimates is strictly smaller than that of joint-sampling PFs [1301.3853][2311.00836].
- **High-dimensional tractability**: Integrating over analytically tractable states avoids the exponential particle-count scaling with total state dimension required by vanilla particle filters.
- **Robustness in the presence of non-Gaussian noise or state constraints**: Hierarchical and constrained RBPFs handle stochastic volatility, heavy-tailed errors [1504.05723], and inequality-constrained estimation [2310.04637].

In POMDPs, RBPF–quadrature hybrids further reduce the number of tree searches required for high-quality planning, outperforming SIR-PF-based planners under fixed computational budgets [2409.16392].

## 5. Methodological Extensions and Pseudocode

RBPF has been widely extended and adapted:
- **Generalization to arbitrary tractable subspaces**: Any subblock for which an optimal, finite-sufficient-statistics filter exists can be marginalized [1301.3853][1707.01311].
- **Look-ahead and decentralized variants**: When exact analytic marginalization is impossible, decentralized filters approximate the Rao–Blackwell step via nested (local) Monte Carlo, with parallelizability and robust adaptive state-space decomposition [1203.2394].
- **Rao–Blackwellized POMDP solvers**: Combine particle-based belief tracking in the hard subspace with analytic filters/UKFs and sparse-grid quadrature for tractable parts, boosting belief accuracy and planning efficiency [2409.16392].
- **Robustification and constraint handling**: RBPFs have been extended to enforce contact constraints via quadratic programming in physical estimation [2310.04637].

A representative high-level pseudocode, tailored to the generic RBPF, is:

```python
for t in 1..T:
    # Step 1: PF prediction for nonlinear/sample-based substate
    for i = 1..N_particles:
        x_tn[i] = propagate(x_{t-1}^n[i], analytic_estimate(x_{t-1}^l[i]))
        # Step 2: Analytic update (e.g., Kalman/HMM) for each particle
        update analytic filter for x_t^l[i] given x_tn[i], y_t
    # Step 3: Compute or integrate observation likelihood for each particle
        w_t[i] = w_{t-1}[i] * likelihood(y_t | x_tn[i], analytic_state[i])
    # Step 4: Normalize and resample if necessary
        normalize w_t
    # Step 5: (Optional) Apply constraint enforcement or specialized updates
```

For application-specific pseudocode, see vehicular GNSS (Algorithm block, [2506.03537]), connected vehicles [1702.05792], superelliptical tracking [2406.10389], and object pose tracking [1905.09304].

## 6. Practical Guidelines, Limitations, and Future Directions

Practical deployment of RBPF requires careful state partitioning and tuning:
- **Partitioning**: Select the largest-possible subspace for analytic marginalization without violating model tractability or introducing significant nonlinearities (e.g., velocity in GNSS or parameters in SDE models [2311.00836]).
- **Noise and model parameter tuning**: Process noise and measurement noise covariance choices are critical for stable analytic filters and reliable particle diversity [2506.03537].
- **Constraint handling**: For systems with physical or combinatorial constraints (contacts, reachability), explicit constraint enforcement per particle via optimization (e.g., QP projection) is necessary [2310.04637].
- **Resampling criteria**: Use effective sample size (ESS)-based thresholds to prevent weight degeneracy and particle impoverishment [1301.3853][1312.0781].

**Current limitations** include:
- Requirement of analytic (finite-dimensional) filters for marginalized subspaces (not all models admit this, though extensions to nested Monte Carlo exist [1203.2394]).
- In models with strong nonlinearities or abrupt switching, state partitioning and filter stabilization can become challenging.
- Computational overhead per particle increases due to analytic updates but is offset by reduced sampling variance and particle count.

**Future extensions** highlighted in recent literature involve:
- Tightly coupled measurement handling (e.g., direct Doppler in GNSS [2506.03537]);
- Modularized parameter-state RBPF with grid-based updating [2311.00836];
- Generalization to arbitrary DBN and graphical model structures [1301.3853];
- Hybrid MCMC–SMC schemes and particle rejuvenation for reduced degeneracy in smoothing [1707.01311][1011.2153];
- Efficient quadrature integration and importance distribution design in high-dimensional planning [2409.16392].

RBPF thus provides a unifying, variance-reducing inference principle across numerous applications in navigation, tracking, hybrid systems, and time-series analysis, with demonstrated advantages in efficiency, robustness, and estimator precision.

Source: https://www.emergentmind.com/topics/rao-blackwellised-particle-filtering-rbpf