Two-Level Fuzzy Control of Independent-Load Hydraulic Systems for Pressure-Flow Composite Regulation in Hydraulic Explosive Bonding
1. Definition and Fundamental Principles
1.1 Independent-Load Hydraulic Systems
An independent-load hydraulic system is a closed-loop pressure-and-flow control architecture in which each actuator or working circuit operates autonomously, decoupled from the primary pump supply. In the context of hydraulic explosive bonding (HEB) and hydraulic cladding, such systems are deployed to deliver precisely controlled, independently actuated pressure pulses to individual bonding zones or workpiece sections. The independence of each load circuit ensures that pressure transients in one zone do not propagate to adjacent zones, which is critical when bonding large-diameter pipes, multi-segment plates, or complex geometries requiring sequential or staggered pressure application.
1.2 Pressure-Flow Composite Control
Pressure-flow composite control refers to the simultaneous regulation of both hydraulic pressure (P) and volumetric flow rate (Q) within a single actuator circuit. In hydraulic bonding operations, pressure governs the magnitude of the bonding force applied to the interface, while flow rate governs the speed and uniformity of the pressure ramp. A composite controller must manage both variables concurrently to achieve the desired bonding trajectory — for example, a slow, uniform pressure rise during the initial contact phase followed by a rapid peak pressure pulse during the explosive bonding event.
1.3 Two-Level Fuzzy Control Architecture
The two-level fuzzy controller architecture described in the referenced study employs a hierarchical fuzzy logic structure:
- Outer Level (Supervisory Controller): A higher-order fuzzy inference engine that evaluates the overall system state — including setpoint tracking error, rate of change of error, and process disturbance indicators — and adjusts the reference parameters of the inner controller in real time. This level handles non-linearities, parameter variations, and external load changes.
- Inner Level (Regulatory Controller): A lower-order fuzzy inference engine that executes fine-grained control of the proportional valves, servo valves, or variable-displacement pumps based on the supervisory setpoints. This level ensures fast response to pressure and flow deviations.
The two-level architecture resolves the inherent trade-off between robustness (outer loop) and responsiveness (inner loop) that plagues single-level fuzzy controllers in hydraulic systems subject to significant non-linearities such as fluid compressibility, valve hysteresis, and variable back-pressure.
2. Category and Business Positioning
2.1 Technical Classification
This research entry falls under the category of Process Control and Automation Engineering supporting the company's hydraulic explosive bonding technology route. It does not directly describe a cladding or bonding process but rather the control infrastructure that enables repeatable, certified hydraulic bonding operations. In the company's capability taxonomy, it is classified as a supporting technology enabling the hydraulic bonding business line.
2.2 Strategic Positioning
The mastery of advanced hydraulic control systems positions the company as a technology-driven manufacturer rather than a process-dependent fabricator. Key strategic advantages include:
- Process repeatability: Fuzzy-controlled hydraulic systems reduce operator dependence and batch-to-batch variability, directly improving first-pass yield rates.
- Scalability: The independent-load architecture allows the same control platform to be scaled from small-diameter pipe bonding (DN50–DN200) to large plate bonding (up to 3000 mm × 3000 mm) without redesigning the hydraulic circuit.
- IP accumulation: The published research (referenced in English) establishes intellectual property and technical credibility in international markets, supporting qualification bids with overseas customers and joint ventures.
- Cross-route applicability: While primarily serving hydraulic bonding, the control principles extend to explosion welding charge placement systems and automated TIG/MIG weld overlay positioning systems.
3. Technical Purpose and Value
3.1 Solving Core Hydraulic Bonding Challenges
Hydraulic explosive bonding requires the application of a precisely shaped pressure waveform to the interface between the base metal and the cladding layer. Conventional PID-based hydraulic controllers struggle with three fundamental challenges:
- Non-linear valve characteristics: Servo and proportional valves exhibit gain variations, hysteresis, and dead-band effects that cause PID controllers to oscillate or lag under transient conditions.
- Variable system compliance: The effective hydraulic stiffness of the bonding assembly changes as the workpiece deforms plastically, altering the pressure-flow relationship dynamically.
- Multi-zone coordination: When bonding a large workpiece with multiple independent hydraulic cylinders, the timing and amplitude of each zone must be coordinated within tight tolerances (typically ±2% pressure accuracy, ±5 ms timing synchronization).
The two-level fuzzy controller addresses all three challenges by providing adaptive gain scheduling (outer loop) and fast disturbance rejection (inner loop) without requiring an exact mathematical model of the hydraulic system.
3.2 Quantified Value Metrics
| Performance Parameter | Conventional PID Control | Two-Level Fuzzy Control | Improvement |
|---|---|---|---|
| Pressure steady-state accuracy | ±5–8% | ±1.5–2.5% | 3–4× better |
| Pressure rise time (0–90% setpoint) | 120–200 ms | 40–80 ms | 2–3× faster |
| Pressure overshoot | 8–15% | 2–5% | Significant reduction |
| Flow rate stability (±1% deviation band) | 60–70% of cycle | 92–97% of cycle | Substantially improved |
| Multi-zone synchronization (max deviation) | 10–25 ms | 2–5 ms | 3–5× tighter |
| Operator intervention frequency | Every 3–5 bonds | Every 20–30 bonds | 5–10× reduction |
4. Key Process and Implementation Points
4.1 System Architecture
The hydraulic control system for bonding operations comprises the following functional layers:
- Power Unit: Variable-displacement axial piston pump with pressure-compensated flow control, typically rated at 21–40 MPa maximum system pressure and 63–250 L/min flow capacity depending on workpiece size.
- Distribution and Isolation: Independent-load manifold block with pressure-relief valves, check valves, and accumulator pre-charges for each bonding zone. Accumulators (typically nitrogen-charged bladder type, 2–10 L volume) provide peak power delivery during the explosive bonding pulse.
- Actuation: Servo-controlled proportional pressure-reducing valves (PRV) and proportional flow-control valves (FCV) at each zone outlet. Valve response time must be ≤10 ms for effective inner-loop control.
- Sensing: High-frequency pressure transducers (0–40 MPa, accuracy ±0.25% FS, response frequency ≥1 kHz) and ultrasonic flow meters at each independent load circuit.
- Control: Industrial PLC or dedicated motion controller running the two-level fuzzy inference algorithm at a cycle time of 1–5 ms. The fuzzy rule base is implemented as a lookup table with interpolation for real-time execution.
- HMI and Data Acquisition: Supervisory SCADA interface for setpoint programming, cycle monitoring, and data logging for traceability.
4.2 Fuzzy Controller Design Parameters
| Parameter | Outer Level (Supervisory) | Inner Level (Regulatory) |
|---|---|---|
| Input variables | Error (E), rate of change of error (ΔE), process disturbance index (D) | Error (e), rate of change of error (Δe) |
| Fuzzification | 5–7 linguistic sets (NB, NM, NS, Z, PS, PM, PB) | 5–7 linguistic sets with triangular membership functions |
| Rule base size | 35–125 rules (3D rule surface) | 25–49 rules (2D rule surface) |
| Inference method | Mamdani with centroid defuzzification | Sugeno (TSK) with weighted average defuzzification |
| Output variables | Adjustment factors for inner-loop gain (Kp_adj, Ki_adj, Kd_adj) | Valve command signal (0–100% normalized) |
| Control cycle time | 10–50 ms | 1–5 ms |
| Adaptation mechanism | Online rule weight adjustment based on performance index | Fixed rule base with gain scheduling from outer loop |
4.3 Pressure-Flow Composite Control Strategy
The composite control strategy operates through three coordinated phases within each bonding cycle:
- Pre-bonding pressurization phase: The inner controller ramps pressure to a holding level (typically 30–60% of peak bonding pressure) while maintaining a controlled flow rate to ensure uniform pressure distribution across the interface. The outer controller monitors for pressure gradient deviations across zones and adjusts individual zone setpoints to compensate for workpiece stiffness variations.
- Peak bonding pulse phase: Upon trigger signal (manual or automatic), the system releases the accumulator charge through the servo valves in a precisely timed sequence. The inner controller modulates valve opening rate to shape the pressure pulse waveform — typically a triangular or trapezoidal profile with a rise time of 20–50 ms, peak hold of 10–30 ms, and decay time of 50–150 ms.
- Post-bonding hold and release phase: Pressure is held at a reduced level (10–20% of peak) for a specified dwell time (500–2000 ms) to allow plastic deformation stabilization, then released gradually to prevent hydraulic shock. The outer controller verifies that all zones have reached the target pressure within the synchronization tolerance before initiating the hold phase.
4.4 Implementation Considerations for Hydraulic Explosive Bonding
- Accumulator sizing: The accumulator must be sized to deliver the peak bonding energy (E = P_peak × V_displacement) within the required pulse duration. For a 20 MPa peak pressure and 5 L effective displacement volume, the minimum accumulator pre-charge energy is approximately 100 kJ.
- Valve bandwidth matching: The servo valve natural frequency must be at least 5× the hydraulic system natural frequency to avoid resonance-induced pressure oscillations. Typical valve bandwidth requirement: ≥50 Hz.
- Fluid condition: Hydraulic oil must maintain viscosity within ISO VG 32–46 range at operating temperature (20–55°C) and cleanliness at NAS 1638 Class 8 or better to prevent valve stiction and sensor contamination.
- Anti-cavitation measures: Minimum line pressure must be maintained above the fluid's vapor pressure at operating temperature (typically ≥0.5 MPa minimum) to prevent cavitation-induced valve damage and pressure measurement errors.
5. Applicable Standards and Acceptance Criteria
5.1 Hydraulic System Design and Testing Standards
| Standard | Scope | Application to Bonding Control System |
|---|---|---|
| ISO 4413 | Hydraulic fluid power — General rules and safety requirements | System design, safety valve sizing, pressure vessel requirements |
| ISO 4414 | Hydraulic fluid power — Designation and dimensions | Component selection, connection standards, nominal size designation |
| ISO 4401 | Hydraulic fluid power — Symbol identification and application | Schematic documentation, valve and circuit symbol standardization |
| ISO 11901 | Hydraulic fluid power — Contamination control | Fluid cleanliness requirements, filtration specifications |
| ISO 4406 | Hydraulic fluid power — Fluid cleanliness code | NAS 1638 / ISO 4406 cleanliness acceptance criteria |
| GB/T 3766 | Hydraulic fluid power — General rules | Domestic compliance for Chinese market hydraulic systems |
| GB/T 19031 | Hydraulic fluid power — Graphical symbols | Standardized hydraulic circuit documentation for Chinese projects |
5.2 Acceptance Criteria for Bonding Control System Performance
- Pressure accuracy: Steady-state pressure deviation from setpoint must not exceed ±2% of full scale across the operating range (5–40 MPa). Verified by calibrated dead-weight tester or reference pressure transducer (accuracy ≤±0.05% FS).
- Pressure rise time: Time from 10% to 90% of peak pressure setpoint must not exceed the WPS-specified value (typically 40–100 ms for hydraulic explosive bonding). Measured with ≥1 kHz sampling rate pressure transducer.
- Multi-zone synchronization: Maximum time deviation between any two zones reaching 90% of peak pressure must not exceed ±5 ms. Verified with synchronized high-speed data acquisition (≥10 kHz sampling rate).
- Cycle-to-cycle repeatability: Peak pressure variation across 20 consecutive cycles must not exceed ±3% of the nominal peak pressure. Standard deviation of the pressure waveform shape factor must be ≤2%.
- Flow rate stability: Flow rate deviation from setpoint during the pressure ramp phase must not exceed ±5% of the setpoint value for any duration exceeding 50 ms.
- System response to load disturbance: Recovery time to within ±2% of setpoint after a sudden 50% load change must not exceed 200 ms.
5.3 Cladding and Bonding Process Standards (Interface)
The hydraulic control system must deliver bonding parameters that satisfy the applicable cladding product standards. The control system's pressure waveform and dwell time parameters are validated against the following standards depending on the application:
- GB/T 25672 — Steel and nickel-clad plate (covers hydraulic bonding methods)
- NB/T 47015 — Technical requirements for welding procedure, welder qualification, and welding management of pressure vessels (relevant when bonded plates are used in pressure vessel fabrication)
- ASTM A270 — Standard specification for clad steel plate and strip for pressure vessels (where ASTM-based qualification is required)
- ASME BPV Section II Part D — Qualification of welders, welding operators, and brazers (when bonded components are incorporated into ASME-coded pressure vessels)
- API 570 — Piping Inspection Code (for bonded piping components in service inspection qualification)
- ISO 9966 — Explosion welding — General guidelines (international reference for explosive bonding process qualification)
6. Common Risks and Controls
6.1 Control System Risks
| Risk Category | Description | Consequence | Mitigation Control |
|---|---|---|---|
| Fuzzy rule base inadequacy | Rule base does not cover all operating conditions, especially at extreme pressures or unusual workpiece geometries | Poor pressure regulation, bonding defects, rework | Offline simulation and rule base validation across the full operating envelope before deployment; online adaptation mechanism with bounded adjustment limits |
| Sensor drift or failure | Pressure transducer or flow meter develops offset or fails | Uncontrolled pressure application, potential over-pressure or under-bonding | Redundant pressure measurement (dual transducers per critical zone); automated calibration check at cycle start; alarm and shutdown on sensor discrepancy >5% |
| Hydraulic fluid degradation | Oil viscosity changes, contamination, or aeration | Valve performance degradation, pressure instability, component wear | Online viscosity monitoring; particle count monitoring; scheduled fluid replacement per ISO 11901; breather filter maintenance |
| Multi-zone asynchronization | One or more zones fail to synchronize with the master zone | Non-uniform bonding, interfacial defects, potential delamination | Zone-by-zone synchronization verification before peak pulse; automatic cycle abort if synchronization tolerance exceeded; root cause analysis of valve or accumulator degradation |
| Accumulator pre-charge loss | Nitrogen charge pressure drops below minimum required level | Inadequate peak pressure delivery, incomplete bonding | Pre-cycle accumulator pressure check; automatic nitrogen recharge system; alarm on pre-charge pressure below 80% of nominal |
| Cybersecurity vulnerability | PLC or SCADA system compromised, leading to unauthorized parameter modification | Equipment damage, safety incidents, product non-conformance | Network isolation of control systems; role-based access control; parameter change audit logging; periodic security audits per IEC 62443 |
6.2 Process Integration Risks
- Workpiece surface condition variability: Surface roughness, contamination, or dimensional tolerance variations between the base and cladding layers affect the pressure-flow relationship. Control: Pre-bonding surface characterization (roughness measurement per ISO 4287, cleanliness verification) and adaptive setpoint adjustment based on measured surface properties.
- Temperature effects on hydraulic performance: Hydraulic oil viscosity changes with temperature, affecting flow rates and valve response. Control: Oil temperature monitoring and compensation; fuzzy controller gain adjustment based on measured oil temperature; heated oil reservoir for cold-start conditions.
- Workpiece stiffness variation: Different workpiece geometries and thicknesses present different hydraulic impedances. Control: Pre-programmed parameter sets for each workpiece configuration; outer-loop adaptive gain scheduling based on real-time pressure-flow ratio monitoring.
7. Application Scenarios Across the Company's Three Technology Routes
7.1 Hydraulic Explosive Bonding (Primary Application)
This is the direct and primary application domain. The two-level fuzzy controlled independent-load hydraulic system is the core enabling technology for the company's hydraulic explosive bonding operations. Specific applications include:
- Large-diameter pipe bonding: DN300–DN2000 carbon steel pipes bonded with stainless steel (304L, 316L, 321), nickel alloys (Inconel 625, Hastelloy C-276), or titanium (Grade 2, Grade 5) cladding layers. The independent-load architecture allows multiple hydraulic cylinders to be arranged circumferentially and axially, with the fuzzy controller coordinating pressure application across all zones to achieve uniform bonding.
- Large plate bonding: Plates up to 3000 mm × 3000 mm × 50 mm bonded with corrosion-resistant or wear-resistant cladding layers. Multi-zone hydraulic systems with 4–16 independent circuits ensure uniform pressure distribution across the large bonding area.
- Complex geometry bonding: Flanges, nozzles, and shaped components where the bonding area is non-planar. The fuzzy controller adapts to the varying hydraulic impedance presented by the complex geometry.
- Multi-layer bonding: Sequential bonding of multiple cladding layers (e.g., carbon steel base + stainless steel intermediate + nickel alloy surface) where each layer requires different pressure parameters. The controller manages the transition between bonding cycles with different setpoints.
7.2 Explosion Welding (Secondary Application)
In explosion welding, the two-level fuzzy control principles apply to the automated charge placement and detonation timing systems:
- Charge placement system control: Hydraulic or pneumatic systems that position and compact explosive charges against the workpiece require precise pressure and flow control. The fuzzy controller ensures consistent charge density and placement geometry across multiple production cycles.
- Detonation synchronization: When multiple detonation points are used for large workpieces, the electrical initiation system's timing control can incorporate fuzzy logic to compensate for propagation time variations across the charge geometry.
- Post-explosion inspection automation: Hydraulic systems for automated ultrasonic testing (UT) probe positioning and coupling fluid application benefit from the same pressure-flow composite control architecture.
7.3 TIG/MIG Weld Overlay (Tertiary Application)
The fuzzy control technology extends to automated weld overlay systems in the following ways:
- Wire feed rate control: In automated TIG or MIG weld overlay, the wire feed motor and torch travel speed must be precisely coordinated. A fuzzy controller can adaptively adjust wire feed rate based on real-time current and voltage feedback to maintain consistent weld bead geometry despite variations in gas shielding flow, torch height, or base metal condition.
- Shielding gas flow control: Precise control of shielding gas flow rate (typically 8–20 L/min for TIG, 15–30 L/min for MIG) is critical to prevent porosity and ensure proper arc stability. Fuzzy control of the gas flow regulator maintains optimal flow despite line pressure variations and ambient wind effects.
- Multi-pass overlay sequencing: For multi-pass weld overlay builds (typically 3–8 passes), the fuzzy controller manages the interpass temperature monitoring and cooling time calculation, adjusting the next pass start time based on real-time thermocouple readings to optimize the thermal cycle and minimize residual stress.
- Positioning system control: CNC-positioned welding systems benefit from fuzzy-controlled axis drives that adapt to the varying friction and inertia encountered during multi-pass overlay on curved or irregular workpiece surfaces.
8. Contribution to Qualification Building, Product Delivery, and Customer Value
8.1 Qualification Building
- WPS/PQR support: The precise and repeatable pressure control enabled by the fuzzy controller directly supports welding procedure qualification (WPS) and procedure qualification record (PQR) development. When hydraulic bonding parameters are required to be within tight tolerances (e.g., peak pressure ±5%, dwell time ±10%), the fuzzy control system provides the capability to meet these requirements consistently, enabling successful qualification testing.
- International certification: The published English-language research demonstrates technical competence and intellectual rigor, which is valued by international certification bodies (e.g., ASME, API, NACE/AMPP) when evaluating the company's manufacturing capabilities. The research publication serves as evidence of the company's investment in process development and quality engineering.
- Customer qualification audits: During customer factory acceptance inspections (FAI) and qualification audits, the ability to demonstrate a sophisticated, well-documented control system with traceable performance data significantly enhances customer confidence. The control system's data logging capability provides objective evidence of process parameter compliance for each production cycle.
8.2 Product Delivery Enhancement
- Reduced scrap rate: Improved pressure control accuracy directly reduces bonding defects such as incomplete bonding, interfacial voids, and delamination. Industry benchmarks suggest that fuzzy-controlled hydraulic systems can reduce scrap rates from 8–15% (conventional control) to 2–4%, representing significant cost savings on high-value alloy cladding materials.
- Increased throughput: Faster pressure rise times and reduced dwell times (enabled by more precise pressure control) increase the number of bonding cycles per hour. A typical improvement is 20–35% throughput increase for hydraulic bonding operations.
- Reduced operator skill requirement: The adaptive fuzzy controller compensates for minor variations in workpiece preparation and environmental conditions, reducing the dependence on highly skilled operators. This enables the company to scale production without proportionally increasing the number of qualified operators.
- Traceability and documentation: Every bonding cycle is automatically logged with complete pressure and flow waveform data, creating a comprehensive traceability record that supports customer quality documentation requirements and facilitates root cause analysis in the event of a field failure.
8.3 Customer Value Creation
- Performance assurance: Customers in critical industries (oil & gas, nuclear power, chemical processing, marine) require assured bonding quality for safety-critical components. The fuzzy-controlled hydraulic system provides quantifiable, verifiable process control that translates into higher confidence in product performance and reduced lifecycle risk.
- Customization capability: The fuzzy controller's adaptability allows the company to offer customized bonding parameters for specific customer applications — for example, optimized pressure waveforms for bonding dissimilar metal combinations with different thermal expansion coefficients, or specialized dwell time profiles for high-temperature service applications.
- Supply chain reliability: Reduced dependence on operator skill and improved process repeatability translate into more reliable delivery schedules and reduced risk of quality-related production delays for the customer's downstream fabrication and assembly operations.
- Technical partnership: The company's demonstrated expertise in advanced process control positions it as a technical partner rather than a commodity supplier. This enables collaborative development of novel bonding solutions, joint qualification efforts, and long-term supply agreements that create sustained value for both parties.
9. Conclusion
The research on two-level fuzzy control of independent-load hydraulic systems represents a foundational capability investment that elevates the company's hydraulic explosive bonding operations from a manually-dependent process to a precisely controlled, data-driven manufacturing system. By integrating adaptive fuzzy logic control into the hydraulic bonding infrastructure, the company achieves quantifiable improvements in bonding quality, production efficiency, and process traceability. These improvements directly support the company's qualification building with international standards bodies, enhance product delivery reliability, and create differentiated value for customers in demanding industrial applications. The cross-route applicability of the control technology — extending to explosion welding charge systems and automated weld overlay platforms — further amplifies the strategic return on this technical investment across the company's entire technology portfolio.