Crane R&D Validation Process: Adams, Simulink and HIL Testing

From Trial-and-Error to Simulation: A New Era in Crane Engineering

In traditional crane development, the "trial-and-error" approach has long dominated—build a prototype, test it, find defects, then rework and start over. For a large smart crane, prototype manufacturing alone can cost millions of RMB, and a single extreme-condition test (such as full-load emergency braking or off-center lifting) risks equipment damage or even safety incidents. As cranes become increasingly intelligent, control system complexity grows exponentially, rendering traditional methods inadequate.

The Kelude R&D team understands this challenge well and has built a "multi-level simulation verification" development framework at the system level. The core philosophy: complete over 90% of control logic validation and over 90% of extreme-condition simulation before physical manufacturing begins, minimizing risks during the real-machine testing phase and cutting development cycles by more than 40%.

Co-Simulation Platform: Adams + Simulink Architecture

Co-simulation is the cornerstone of Kelude's R&D framework. The platform is powered by two core engines:

Adams (Multi-Body Dynamics Simulation)

Adams (Automatic Dynamic Analysis of Mechanical Systems) builds the crane's mechanical model—including all kinematic joints such as the boom, slewing mechanism, hoisting mechanism, travel mechanism, and wire rope pulley system. By precisely defining mass, inertia, stiffness, and damping parameters for each component, Adams delivers high-fidelity simulation of dynamic responses under complex operating conditions, covering structural elastic deformation, multi-body coupled vibration, and nonlinear contact phenomena. The model also incorporates flexible-body characteristics of wire ropes, realistically simulating pendulum dynamics during hoisting operations.

Simulink (Control System Simulation)

Simulink hosts the crane's control algorithm models—including speed/position closed-loop control for each mechanism, anti-sway control logic, load moment limiter (LML) algorithms, and safety interlock logic. The control logic running in the Simulink model remains strictly consistent with the program ultimately deployed to the actual PLC.

Co-Simulation Interface

Adams and Simulink exchange data bidirectionally in real time through the Co-Simulation interface: Adams transmits real-time states—position, velocity, acceleration, torque—of each mechanism to Simulink, while Simulink returns controller output commands (such as VFD frequency setpoints and brake actuation signals) to Adams. This closed-loop simulation architecture enables R&D engineers to execute the complete verification chain—"sensor sampling → controller computation → actuator response → mechanical system dynamics"—entirely within a virtual environment.

SIL (Software-in-the-Loop) Testing Process

SIL testing serves as the first verification gate above co-simulation. During the SIL phase, control software (not yet deployed to PLC hardware) is compiled into a PC-executable program and run on the same or a networked computing node alongside the Adams/Simulink co-simulation model.

The core value of SIL testing lies in:

• Control logic correctness verification: Confirming that speed curves, acceleration/deceleration times, and safety interlock conditions for each mechanism align with design specifications
• Boundary condition testing: Exercising control behavior under all input signal combinations at the software level, covering both normal and abnormal operating sequences
• Regression testing: Automatically running the SIL test suite after every control software iteration to ensure modifications introduce no new defects

Kelude's team has built a library of over 200 SIL test cases, covering all operating modes and safety functions.

HIL (Hardware-in-the-Loop) Testing: Real PLC + Virtual Plant

HIL testing is the pivotal link in Kelude's R&D framework. Its core architecture:

Real PLC (S7-1500) + Real-Time Simulator (running Adams dynamics model in real time)

In HIL testing:

1. The device under test is a real S7-1500 PLC running the final version of the control program (identical to the production machine)
2. The PLC communicates with the real-time simulator via PROFINET/Profinet IO
3. The simulator runs the Adams dynamics model in real time, feeding virtual sensor signals back to the PLC (angle encoders, load sensors, limit switches, etc.)
4. PLC outputs (VFD control words, brake signals, etc.) are received by the simulator in real time to drive the virtual mechanical model

The key advantage of this architecture: PLC hardware in the loop, real control program execution, and virtual plant response—all three simultaneously satisfied, enabling comprehensive hardware-level control system verification without requiring a physical crane.

Extreme-Condition Test Coverage

Kelude's R&D verification framework systematically simulates the following extreme operating conditions:

Each operating condition generates a detailed simulation test report, including time-domain response curves, spectrum analysis, fatigue damage assessment, and other quantitative data, which serve as the baseline reference for physical machine verification.

From Simulation to Physical Testing: A Closed-Loop Verification Workflow

Kelude's R&D verification system follows a complete "V-shaped" validation workflow:

1. Requirements Analysis & Simulation Planning – Define test cases and acceptance standards at each level based on design specifications and code requirements.
2. Model Development & Calibration – Build Adams dynamics models and Simulink control models, then calibrate them using component-level test data.
3. SIL Verification – Validate control logic at the pure software level.
4. HIL Verification – Run real-time simulation with PLC hardware-in-the-loop.
5. Physical Machine Testing – Execute critical operating condition tests on an actual crane and capture real-world measurement data.
6. Data Feedback & Model Iteration – Feed physical test data back into the simulation models, refine parameters, and improve simulation fidelity.

This closed-loop mechanism allows simulation models to continuously "evolve" as project data accumulates, steadily narrowing the gap between simulation and real-world performance. According to Kelude's internal data, after three or more rounds of model iteration and calibration, the deviation between simulation and physical test results for critical operating conditions has been reduced to within 5%.

R&D Team Structure and Patent Portfolio

Kelude's R&D team is organized around a three-pillar architecture: system-level simulation, control algorithms, and hardware testing.

• System Simulation Group – Responsible for Adams multi-body dynamics modeling, co-simulation platform maintenance, and design of extreme operating condition test cases.
• Control Algorithm Group – Handles Simulink control model development, anti-sway algorithms, path planning algorithms, and safety logic design.
• Test & Validation Group – Manages SIL/HIL test platform setup, test execution, and analysis report preparation.

On the patent front, Kelude has filed multiple invention patents centered on its simulation and verification system, covering core areas such as co-simulation interface methods, HIL test apparatus structures, automatic generation algorithms for extreme-condition test matrices, and closed-loop calibration methods between simulation and physical testing. These patents not only protect the company's technical achievements but also establish a strong moat in the field of smart crane simulation and verification.

It's worth noting that Kelude's R&D verification system is not limited to new product development—it is equally applicable to intelligent retrofit projects for cranes already in service. The key difference is that retrofit projects require reverse modeling based on measured structural parameters of the existing equipment, with particular emphasis on verifying interface compatibility and safety interlock logic between the old and new control systems. For special simulation and verification requirements specific to explosion-proof cranes, please refer to the explosion-proof technology section; for real-world applications of simulation verification in retrofit projects, see the retrofit case studies section.

Economic Benefits and Technical Value of the Verification System

From an economic standpoint, the return on investment for Kelude's R&D verification system is substantial. According to data from the R&D center:

• Fewer Prototype Iterations – Reduced from an average of 4–5 physical prototype iterations in the traditional model to just 1–2 today, with some mature models achieving "design once, build once, pass once" on the first attempt.
• Shorter Test Cycles – The complete verification cycle from simulation to physical testing has been compressed from 12–18 months down to 7–10 months.
• Lower Field Failure Rates – On-site control system failure rates within 12 months of commissioning have dropped by 76% compared to pre-system levels.
• Higher Certification Pass Rates – The first-attempt pass rate for type tests has improved from 65% to over 92%.

From a technical perspective, the value of this system manifests at three levels. First, it establishes a digital R&D pipeline from virtual simulation to physical hardware, enabling design changes to be validated through simulation within hours rather than waiting weeks for prototype modifications. Second, the vast amount of simulation data lays a solid foundation for future intelligent upgrades of downstream products, such as digital twin implementations and predictive maintenance. Third, the standardized verification process enables efficient knowledge reuse across different projects, significantly shortening the training period for new engineers.

Kelude's R&D verification system has completed three rounds of internal review and technical iteration, and the company is actively pursuing software copyright registrations and technical standard initiatives. Looking ahead, the team plans to extend the verification system into full life cycle management for cranes, including digital twin-based in-service health monitoring, simulation-driven remaining life prediction, and HIL-based remote software upgrade validation.

Frequently Asked Questions (FAQ)