Disturbance Observer and Optimal LQR Composite Control for Hydro-Electric Systems in Cladding Manufacturing

1. Definition and Technical Principles

The composite control strategy combining a Disturbance Observer (DOB) with an optimal Linear Quadratic Regulator (LQR) represents a sophisticated approach to managing hydro-electric servo systems used in high-precision cladding and bonding equipment. This control architecture addresses the fundamental challenge of maintaining sub-millimeter positioning accuracy and consistent pressure profiles in hydraulic systems that are inherently subject to nonlinearities, parameter variations, and external disturbances.

The core principle operates on two complementary layers:

The composite architecture integrates these two layers such that the LQR handles the nominal system dynamics and setpoint tracking, while the DOB provides adaptive compensation for unmodeled dynamics and external perturbations. The combined system achieves performance that neither controller could deliver independently.

2. Category and Business Positioning

This control technology falls within the category of Advanced Process Automation and Equipment Control Engineering, serving as a critical enabler for the precision and repeatability required in cladding manufacturing. Within the company's technology portfolio, it occupies the infrastructure layer that underpins all three primary technology routes:

The business positioning of this competency is as a competitive differentiator that enables the company to deliver products with superior dimensional accuracy, reduced scrap rates, and shorter qualification cycles compared to competitors relying on conventional PID-based hydraulic controls.

3. Technical Purpose and Value

3.1 Precision Enhancement

Hydraulic systems in cladding equipment are subject to inherent challenges including fluid compressibility, valve hysteresis, seal friction, and thermal expansion. Conventional proportional-integral-derivative (PID) controllers often exhibit oscillation, overshoot, or slow response when faced with these nonlinearities. The DOB-LQR composite control reduces positioning error by 60–80% compared to PID control, achieving steady-state accuracy within ±0.02 mm for hydraulic ram positioning and ±0.5% for pressure control.

3.2 Disturbance Rejection

In hydraulic explosive bonding, the bonding pressure must be held constant for a specified dwell time to ensure metallurgical interface formation. Load disturbances from material deformation, thermal expansion, and pump pulsation can cause pressure drift. The disturbance observer continuously estimates and compensates these disturbances, maintaining pressure stability within ±1% of setpoint throughout the bonding cycle.

3.3 Energy Efficiency

The LQR optimization minimizes unnecessary control effort, reducing hydraulic pump cycling and valve chatter. This translates to lower energy consumption per production cycle, reduced hydraulic fluid degradation, and extended equipment service life.

3.4 Process Qualification Support

For welding procedure qualification under ASME Section IX or ISO 15614, the repeatability of process parameters is a critical acceptance criterion. The composite control system's ability to maintain consistent parameter delivery cycle-to-cycle directly supports qualification testing and reduces the number of test coupons required.

4. Key Process and Implementation Points

4.1 System Architecture

Component Function Key Specifications
Hydraulic Actuator (Ram/Cylinder) Primary force generation Working pressure: 100–4,000+ MPa; Stroke: 50–500 mm
Proportional Servo Valves Flow and pressure modulation Bandwidth: ≥100 Hz; Repeatability: ±0.5% full scale
Position Sensors (LVDT/Laser) Ram position feedback Resolution: ≤1 μm; Linearity: ±0.01% stroke
Pressure Transducers System pressure feedback Range: 0–4,500 MPa; Accuracy: ±0.25% FS
Industrial Controller (PLC/Real-Time PC) DOB + LQR algorithm execution Cycle time: ≤1 ms; OS: Real-time Linux or equivalent
Hydraulic Power Unit Fluid supply and pressure generation Flow: 10–200 L/min; Pressure: up to 4,500 MPa

4.2 Disturbance Observer Design Parameters

Parameter Description Typical Range
Observer bandwidth (ωc) Frequency at which disturbance estimation converges 50–200 rad/s
Plant model order Order of the reference model used for estimation 2nd to 4th order
Filter coefficient Low-pass filter applied to disturbance estimate 100–500 rad/s
Sampling interval Control loop update rate 0.1–1 ms

4.3 LQR Weight Matrix Selection

The performance of the LQR controller is critically dependent on the selection of the state weighting matrix Q and control effort weighting matrix R. The design procedure involves:

  1. Defining the state vector x = [position, velocity, pressure, flow error]T
  2. Assigning higher weights in Q to position and pressure errors to prioritize tracking accuracy
  3. Assigning moderate weights in R to control effort to limit valve saturation and pump cycling
  4. Solving the continuous-time algebraic Riccati equation (CARE): ATPA − PAP + Q − PBR−1BTP = 0
  5. Extracting the optimal feedback gain: K = R−1BTP
  6. Validating closed-loop pole placement and robustness margins

4.4 Implementation Procedure

  1. System Identification: Characterize the hydraulic plant dynamics through frequency response analysis (sine sweep or step response testing) to obtain an accurate transfer function model.
  2. Reference Model Design: Develop the nominal plant model Gn(s) used by the disturbance observer, ensuring it captures dominant dynamics within the operating bandwidth.
  3. LQR Synthesis: Compute optimal gains for the nominal model, selecting Q and R matrices to achieve desired transient response (rise time, settling time, overshoot).
  4. DOB Tuning: Adjust observer bandwidth and filter coefficients to balance disturbance rejection speed against noise sensitivity.
  5. Integration and Validation: Combine DOB and LQR in the control loop, validate through simulation (MATLAB/Simulink) and hardware-in-the-loop testing.
  6. Field Commissioning: Deploy on production equipment, perform parameter fine-tuning under actual operating conditions, and document final control parameters for WPS/qualification records.

5. Applicable Standards and Acceptance Criteria

5.1 Control System Performance Standards

5.2 Equipment and Process Standards

5.3 Acceptance Criteria for Control System Integration

Criterion Acceptance Threshold Test Method
Positioning accuracy (steady-state) ≤ ±0.05 mm Step response test with LVDT measurement
Pressure control stability ≤ ±1% of setpoint over dwell period Continuous pressure monitoring during bonding cycle
Disturbance rejection time ≤ 50 ms for step disturbance Applied load perturbation test
Overshoot (position) ≤ 2% of setpoint Step response measurement
Cycle-to-cycle repeatability ≤ ±0.5% parameter variation Statistical analysis of 30 consecutive cycles
Control loop response time ≤ 1 ms Signal injection and timing measurement

6. Common Risks and Controls

6.1 Model Mismatch Risk

Risk: The disturbance observer relies on a reference plant model. If the actual hydraulic system dynamics deviate significantly from the model (due to fluid contamination, seal wear, temperature changes, or component replacement), the observer may produce incorrect disturbance estimates, leading to degraded performance or instability.

Controls: Implement adaptive observer bandwidth tuning; incorporate periodic system identification routines; use robustness analysis (μ-synthesis or H∞ methods) during design to ensure stability margins under model uncertainty; establish maintenance schedules for hydraulic components that affect plant dynamics.

6.2 High-Frequency Noise Amplification

Risk: The disturbance observer may amplify high-frequency measurement noise, particularly from pressure transducers and position sensors, leading to excessive control signal activity, valve wear, and potential actuator damage.

Controls: Apply appropriate low-pass filtering to sensor signals before observer input; design observer bandwidth below the noise frequency content; implement rate limiting on control signals; use high-quality sensors with adequate noise rejection.

6.3 Parameter Sensitivity in LQR Design

Risk: Poorly chosen Q and R matrices can result in aggressive control action (high Q, low R) causing actuator saturation, or sluggish response (low Q, high R) failing to meet performance specifications.

Controls: Use systematic parameter selection methods (e.g., Bryson's method, eigenvalue assignment); validate through simulation across the full operating envelope; implement gain scheduling for different operating conditions; maintain documented control parameter records for traceability.

6.4 Hydraulic System Nonlinearities

Risk: Coulomb friction, valve deadband, fluid compressibility nonlinearity, and actuator end-of-stroke behavior violate the linear assumptions underlying both LQR and DOB designs.

Controls: Incorporate nonlinear compensation terms in the controller design; use piecewise-linear models for valve characteristics; implement anti-windup strategies for integral terms; design for operation within the linear region of component characteristics.

6.5 Cybersecurity and Data Integrity

Risk: Networked control systems are vulnerable to cyber attacks that could compromise safety-critical cladding operations, potentially leading to product failure or equipment damage.

Controls: Implement IEC 62443-compliant cybersecurity measures; segment control networks from corporate IT; implement data integrity checks on sensor inputs; maintain air-gapped backup control systems for critical operations.

7. Application Scenarios Across Company Technology Routes

7.1 Hydraulic Explosive Bonding (Hydrostatic Extrusion)

In hydrostatic extrusion cladding, the composite control system manages the multi-stage pressure profile applied to the composite billet. The process typically involves:

The DOB-LQR system ensures that pressure ramps are smooth and accurate (avoiding shock loading that could cause cracking), that bonding pressure is held within tight tolerance during the critical dwell period, and that any load disturbance from material deformation is compensated in real-time. This directly impacts the quality of the metallurgical bond and the pass/fail rate of shear test verification per ASTM E2744 or equivalent.

7.2 Explosion Welding

While explosion welding is a high-energy process, the composite control system governs the precision positioning and separation distance control of the flyer and base plates prior to detonation:

The disturbance observer compensates for thermal expansion of fixtures, vibration from nearby operations, and hydraulic drift, ensuring that the separation distance at the moment of detonation matches the design specification. Even a 0.5 mm deviation can change the collision angle by several degrees, potentially shifting the bonding zone from the optimal sinusoidal interface to a non-bonded or fractured condition.

7.3 TIG/MIG Weld Overlay

In robotic weld overlay applications, the composite control system governs the hydraulic or electric actuators that position the welding torch and workpiece:

For multi-layer overlay builds (e.g., 3–5 passes of 309L followed by 316L), the system must maintain consistent parameters across all passes while compensating for cumulative thermal distortion. The disturbance observer estimates and compensates for workpiece deflection, while the LQR ensures smooth, vibration-free torch movement that prevents arc instability and porosity formation.

8. Contribution to Qualification Building and Customer Value

8.1 Welding Procedure Specification (WPS) Qualification

The composite control system directly supports WPS qualification by providing documented evidence of parameter control capability. When submitting a WPS for approval under ASME Section IX, ISO 15614, or NB/T 47014, the following control-related documentation can be provided:

8.2 Product Delivery Quality

For large-scale cladding projects (e.g., pressure vessel linings, heat exchanger tubes, or pipeline overlays), the composite control system enables:

8.3 Customer Value Proposition

The advanced control capability positions the company as a technology leader capable of delivering high-integrity cladding products for demanding applications including:

8.4 Continuous Improvement and Digitalization

The composite control system generates high-frequency process data that can be leveraged for:

9. Conclusion

The integration of disturbance observer and optimal LQR composite control into hydro-electric systems represents a fundamental advancement in the precision manufacturing capabilities of Cladding Technology Shanxi Co., Ltd. This technology transforms conventional hydraulic equipment into intelligent, adaptive manufacturing platforms capable of meeting the stringent quality requirements of nuclear, energy, and chemical processing industries. By ensuring parameter consistency, rejecting process disturbances in real-time, and generating comprehensive process data, the composite control system directly contributes to qualification success, product quality, customer satisfaction, and long-term competitive positioning in the global cladding and overlay manufacturing market.