Disturbance Observer Techniques for Robotic Force Control

Disturbance observers have emerged as a cornerstone in modern robotic force control, offering a sensorless approach to estimating and compensating for dynamic uncertainties and external interactions. By modelling the robot’s nominal dynamics and treating unmodelled effects—such as friction, payload variations and contact forces—as disturbances, these observers reconstruct the unknown inputs from available motor signals. When integrated with force-control architectures such as impedance or admittance controllers, disturbance observers allow robots to achieve compliant, accurate contact tasks without dedicated force-torque sensors. Advances over the past decade have seen the development of linear observers based on momentum or velocity estimation, nonlinear observers exploiting sliding-mode or Kalman-filter principles, and extended state observers that augment the system with additional states for disturbance estimation. More recently, universal nonlinear observers have relaxed design constraints, offering faster convergence and higher accuracy across a wide range of manipulators. Hybrid control schemes that combine adaptive impedance and admittance strategies with real-time disturbance estimation have demonstrated significant reductions in actuator effort while preserving stability under variable interaction conditions. These techniques underpin applications from exoskeleton rehabilitation and collaborative industrial robots to hazardous-environment manipulators, enhancing safety, reducing hardware costs and broadening the scope of force-sensitive automation.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent studies have showcased the continual refinement and diversification of observer-based force-control methods. A novel hybrid adaptive control scheme for upper-limb rehabilitation robots employs an extended state observer to estimate interaction torque without force sensors. This approach updates impedance gains in real time according to patient contact stiffness, yielding markedly lower actuator torques compared with conventional hybrid frameworks. In parallel, a universal nonlinear disturbance observer has been proposed to overcome restrictive assumptions of earlier designs. The new observer ensures bounded estimation error and uniformly faster disturbance tracking, as validated through simulation on multi-degree-of-freedom manipulators. Complementing these design-focused advances, a survey of five representative observer classes—including generalised momentum, velocity-based, extended state, disturbance Kalman filter and nonlinear variants—compares their performance under constant and time-varying disturbances. The survey highlights trade-offs in implementation complexity, disturbance bandwidth and noise sensitivity, providing practical guidance for controller integration in diverse robotic applications.

Disturbance Observer Techniques for Robotic Force Control publication trend

The graph below shows the total number of articles in disturbance observer techniques for robotic force control across all publications each year (not limited to Nature Index journals).

Technical terms

Disturbance observer: A computational scheme that estimates unmeasured external forces or model uncertainties by comparing actual system outputs with outputs predicted by a nominal dynamic model.

Impedance control: A force-control strategy that regulates the dynamic relationship between contact force and motion by shaping the robot’s apparent mechanical impedance (stiffness, damping, inertia).

Admittance control: A complementary strategy that computes desired motion in response to measured or estimated forces, effectively controlling the robot’s apparent compliance.

Extended state observer: A linear observer that augments the system state with an additional variable representing lumped disturbances, facilitating real-time disturbance estimation and compensation.

Nonlinear disturbance observer: An observer design that uses nonlinear estimation techniques—such as sliding-mode or adaptive laws—to achieve robust disturbance tracking in the presence of model uncertainties and variable interaction forces.

References

1. Hybrid Adaptive Impedance and Admittance Control Based on the Sensorless Estimation of Interaction Joint Torque for Exoskeletons: A Case Study of an Upper Limb Rehabilitation Robot. Journal of Sensor and Actuator Networks (2024).
2. Universal nonlinear disturbance observer for robotic manipulators. International Journal of Advanced Robotic Systems (2023).
3. A brief survey of observers for disturbance estimation and compensation. Robotica (2023).