Digital Twins for Permissible Noise Level

Challenge

As more transport projects are planned to meet growing mobility needs, the tension between gaining transport benefits and local environmental impacts inevitably rise. The benefits of greater transport capacity and faster journey times are often offset by the disruption caused during the construction of the new infrastructure.

With rising costs and a need to balance environmental impact with greater mobility, the HS2 Innovation Team wanted to investigate how to maximise train throughput without exceeding the permissible noise levels. To reduce reliance on expensive and time-consuming experimental studies, HS2 worked with CFMS to use digital engineering tools that have been developed in aerospace and other high value design sectors. Transferring these tools from other industries provided actionable insight to optimise the timetable schedule while checking that noise limits would not be exceeded

Solution

Modern simulation techniques coupled with multi-objective optimisation and sensitivity analysis provide a means to better understand the uncertainty in any design. Using a combination of model-based engineering, advanced simulation and high-performance computing, we built a software model and simulated different scenarios of different train volumes passing along the network.

Using a combination of noise models, geographic data and system constraints, a digital twin was created of a section of the track in northwest London, with 166 surrounding assessment locations. The digital twin enabled the direct assessment of resulting noise levels (DARN) to be calculated at each location of this section. By modelling different train volumes using the track that did not exceed the permitted noise level limits, the optimised frequency of trains was explored.

A range of methods were used to determine which assessment locations exceeded noise limits and at what train frequencies. Combined with an in-house framework for optimisation and sensitivity analysis, this gave HS2 planners an ideal throughput of trains and confidence in the forecasts.

Impact

With a digital twin to model different scenarios, this approach can be readily used to automatically optimise train system throughput for a range of different starting scenarios such as asset failure, extreme weather or maintenance. The project highlights how digital engineering methods can quickly generate accurate models, improving infrastructure projects.

These tools can also aid at the conceptual design stage. Key performance metrics can be quantified from simulation results and used to assess whether a given design satisfies requirements, goals and constraints. Multi-objective optimisation techniques can be applied to this complete system model to provide a range of optimal designs that exhibit different trade-offs between goals, enabling for better understanding of the impact of design decisions.