Education
Inner Mongolia University of Science and Technology
MEng · Control Engineering
Research and Engineering
Xiaopei Wang
Researcher · Engineer
PhD Candidate in Electrical Power and Energy · Universiti Malaya
Co-founder and Chief Engineer · Qingrun Hengjie
Complex Energy and Automotive Powertrain SystemsIntelligent Testing
Virtual–Physical Validation
and Engineering Optimisation
xiaopeiwang.com ↗Virtual–Physical Validation
and Engineering Optimisation
Background
Education, Career and Academic Exchange
Education, R&D and academic exchange on a shared timeline.
Institution and Role
Period201820202022202420252026.10
Education
Education
Universiti Malaya
PhD Candidate · Electrical Power and Energy
Career
Career
Tsinghua University
Lead, Hydrogen and Fuel Cell Testing
Career
Huairou National Laboratory
Researcher and Engineer
Career
Qingrun Hengjie
Co-founder and Chief Engineer
Visits & Exchange
Visits & Exchange
National Taiwan University
Visiting PhD Student · Carbon Accounting
Visits & Exchange
The University of Tokyo
Visiting PhD Student · Energy Systems Modelling
Visits & Exchange
Hanoi University of Science and Technology
Visiting Researcher · Energy Management Control
Research Overview
Research Themes and Supporting Methods
Two application areas define the questions; shared methods and engineering feedback advance the work.
Research · Hydrogen and Fuel Cells
Fuel Cells: Controlled Dynamic Testing
Studying test conditions and stack response through gas supply, pressure, humidity, temperature and load.
Project Basis
Development and delivery of a 1 kW fuel cell test benchCore engineer contributing to the design, control development and validation of the 1 kW fuel cell test bench.
Research · Hydrogen and Fuel Cells
Auxiliary Systems: Test Boundary Control
Gas supply and thermal management together shape the test environment.
Project Basis
Gas supply control R&D; delivery of a thermal management subsystem test benchThese projects supported parallel-MFC accuracy optimisation (Processes, 2026) and thermal management testing (GSETS, 2025), respectively.
Research · Hydrogen Operations
Hydrogen Value Chain: Monitoring and Analysis
Studying hydrogen production, storage, transport, refuelling and vehicle operations as a connected system.
Project Basis
National Key R&D Programme · Science and Technology for the Winter Olympics: Hydrogen Mobility R&D and Demonstration2020–2023 · Core engineer contributing to the hydrogen data platform and demonstration analysis; related outputs include 7 public datasets.
Research · Vehicle Powertrains
Powertrains: Loading and Performance Testing
Repeatable loading and measurement for motors, electric drives, hybrid powertrains and coupled shafts.
Project Basis
Delivery of hybrid powertrain test benches for a leading new energy vehicle manufacturerProject Lead; responsible for system design, control architecture and core software, with 22 production-ready systems delivered.
Research · Shared Methods · Virtual–Physical Testing
X-in-the-Loop Validation
Connecting models, software and physical systems for feedback tests that answer specific validation questions.
Define Tests→Execute and Measure→Record and AnalyseFeed results back into models, parameters and test cases
Applications
Fuel cell control tests, powertrain testing and general SIL/HIL platforms, covering models, interfaces, scenarios and results.
Software Outputs
Test control software (2025SR0697142, first contributor); fuel cell model library (2024SR0302142, second contributor).
Case 01 · Fuel Cell Testing and Control Development · 01 / 04
Coordinated Fuel Cell Control and Layered Testing
Fuel cells convert hydrogen into electricity through coordinated air supply, hydrogen supply, cooling and electrical control.
Air supply | Hydrogen recirculation | Cooling | Electrical control
Challenge: coordinated component response
Air, hydrogen and cooling must follow power demand. Startup, operation and shutdown also require coordinated actions.
Approach: introduce physical hardware in stages
Check control logic, connect the physical controller, then test components or the stack to assess interfaces and physical responses.
Purpose: real components with adequate instrumentation
Retain vehicle component characteristics while providing measurement, control and recording for integration and calibration.
Platform value: observable, recorded control actions and physical responses feed back into system development.
Case 01 · Fuel Cell Testing and Control Development · 02 / 04
Models, Controllers and Physical Tests
Each layer has a defined test object; measured data feeds back into control development and calibration.
Models and Control
Virtual plant + control (MIL)
Tech Plant and control models
Use Check logic and dynamics
Models / conditions
Control software
Control software
Controller HIL
Physical controller + real-time plant model
Tech Real-time model + CAN/I/O
Use Check I/O and responses
Interface definitions
Control commands
Control commands
Physical Testing
Choose CIL or stack testing to suit the task
80 kW CIL
Real air, hydrogen recirculation and cooling parts
Tech Parts + instrumentation
Use Calibrate and integrate
80 kW Stack Bench
Real stack with gas supply, cooling and load
Tech Gas supply, cooling and load
Use Enable stack tests
↑ Model and control updates← Calibration and test feedbackMeasured data ↓
Case 01 · Fuel Cell Testing and Control Development · 03 / 04
Implementing the 80 kW CIL Platform
Sensors measure states, the controller commands components, and supervisory software manages operation, monitoring and recording.
Measurements→Control→Valve / pump actions↩ Feedback
Component integration
Integrate air supply, hydrogen recirculation, cooling and instrumentation.
Signals and feedback
Feed temperature, pressure and flow to the controller; command the components.
Operating states
Sequence initialisation, power-up/down, operation, purging and shutdown.
Built the platform, supervisory software and acquisition interfaces; commissioned air and cooling subsystems.
Engineering role: Project lead responsible for the overall design, virtual–physical test architecture and core control-system development; contributed to integration and onsite commissioning.
Case 01 · Fuel Cell Testing and Control Development · 04 / 04
Control Validation and Calibration
Check action sequences and physical responses; use observed differences to refine controls and parameters.
High-voltage precharge: from logic to hardware
Precharge before closing the main circuit to reduce inrush current
Result Models check timing; bench tests record relay commands, power-up states and compressor voltage responses.
Air supply: measurements guide calibration
Adjust target flow and pressure with load demand; control the compressor and back-pressure valve
Result Recorded responses support control calibration and recalibration of valve command–feedback relationships.
Engineering value: testable hardware actions, measured responses for calibration, and a platform and software foundation for system development.
Case 02 · Beijing 2022 Hydrogen Supply Chain Data Platform · 01 / 04
Beijing 2022 Deployment and Research Use
Deployed at the Zhangjiakou Hydrogen Energy Research Institute to monitor hydrogen mobility during the Winter Games.
117
buses
25
hydrogen transport trucks
2
hydrogen refuelling stations
1 hydrogen plant
production equipment monitored
Application: operating states, analysis and alerts
Monitor operations and equipment, analyse performance and safety, and display results and fault alerts at the institute and vehicle manufacturer.
Continued value: evaluation and research
Operating records support service, safety and benefit evaluation. Software, public datasets and evaluation outputs support further hydrogen research.
Supply-chain data supported Games-time operations and subsequent demonstration evaluation and research.
Case 02 · Beijing 2022 Hydrogen Supply Chain Data Platform · 02 / 04
Monitoring the Hydrogen Supply Chain
Hydrogen vehicles need a continuous supply. Production, storage, transport, refuelling and use must be monitored together.
Challenge: multiple supply-chain operators
Vehicles, stations, transport fleets and production equipment have separate data systems; one link cannot show the whole chain.
Production→Storage and transport→Refuelling→Vehicle use
Objective: a shared view of operations
Connect enterprise and equipment data, consolidate operating states, and support monitoring, alerts and analysis.
Purpose: operational support and evaluation
Operators track supply and demand; researchers and managers use recorded data to evaluate the demonstration.
Platform value: turn dispersed records into supply-chain operating information and analysis for hydrogen mobility.
Case 02 · Beijing 2022 Hydrogen Supply Chain Data Platform · 03 / 04
Layered Processing of Hydrogen Data
Data ingestion, organisation and layered processing turn equipment records into information for monitoring and analysis.
① Multi-source ingestion
Connect vehicles, hydrogen transport vehicles, stations and production equipment; decode their messages.
Output: connected data channels
→
② Data organisation
Register equipment and business information; store static, dynamic and display data in separate layers.
Output: queryable operating records
→
③ Layered processing
Rapid processing detects anomalies, real-time processing computes operating metrics, and offline analysis examines historical records.
Output: alerts, metrics and analysis
→
④ Monitoring and analysis
Display supply-chain states and alerts; use analysis to evaluate the demonstration.
Output: monitoring and evaluation data
Why layered processing?
Anomaly detection needs prompt feedback; statistics and more complex analyses use separate processing paths.
From records to analysis
Clean and organise samples for each task, then convert speed, distance and hydrogen records into operating metrics.
Case 02 · Beijing 2022 Hydrogen Supply Chain Data Platform · 04 / 04
Supply-Chain Monitoring and Transport Alerts
A shared interface shows supply-chain states; dedicated functions turn location and operating records into monitoring information.
Supply-chain monitoring: supply and use
Equipment data → States and metrics → Shared display
Combine production, transport, refuelling and vehicle information to review operations, alerts and performance analysis.
Transport monitoring: records inform alerts
Location and operating data → Route / geofence checks → Alerts
Combine routes, geofences and vehicle information to monitor hydrogen transport and identify anomalies.
Engineering role:Core engineer responsible for platform development and deployment, monitoring metrics, data analysis and data submission.
Technology · Framework
Engineering AI and Test Control
Building on test control, data and optimisation methods to integrate engineering AI with test systems.
Models · Equipment · Test ControlEngineering and research foundations
Engineering AI Development
Research and adaptation directions
Engineering Data
Operating data and knowledge
Local GPU
Local LLM deployment
QwenLoRA · Planned
Model Adaptation
Task and domain adaptation
Engineering Support
Knowledge · Test Cases · Reports
Established Methods and Test Control
Data—Models—Optimisation—ExecutionReliable Sensing
Acquisition and quality assessment
State Inference
Modelling and window prediction
Optimisation
Targets, constraints and parameters
Test Execution
Control, interlocks and loads
Virtual–Physical ValidationMIL / HIL / CIL / Hardware→Measurement Feedback↺Data, Model and Parameter Refinement
Integration framework: engineering AI provides assistance; test control handles real-time control and equipment protection.