Virtual Shock-Load Sensor Using Digital Twin
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for measuring shock loads in vehicles are costly and impractical for production vehicles, and there is a need for a cost-effective and time-efficient method to infer shock loads and predict component failures without requiring actual sensors or extensive data storage and transfer.
Innovation Solution
A method using a virtual sensor that infers shock loads from data available on a vehicle's Controller Area Network (CAN) by creating a digital twin and simulating vehicle motion to derive a relationship between CAN signals and shock loads, allowing for continuous monitoring and prediction of component failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional sensors (strain gauges, accelerometers) are mounted on the vehicle to directly measure shock loads, then measurement precision is improved, but device complexity and cost increase significantly
Solution Approach 1:
The patent creates a digital twin (virtual copy) of the vehicle that includes virtual sensors replicating the functionality of physical sensors. This digital twin is trained using data from vehicles equipped with additional sensors, allowing the virtual model to predict shock loads without requiring physical sensor installation on production vehicles.
Solution Approach 2:
The patent replaces the mechanical sensor system with a computational model. Instead of using physical strain gauges and accelerometers to measure shock loads, the system uses a trained neural network that processes data from existing vehicle sensors (CAN bus data) to infer shock load information.
2Reliability
If additional sensors are installed on production vehicles to measure shock loads, then reliability of shock load detection is improved, but manufacturing cost increases
Solution Approach 1:
The patent makes existing vehicle sensors serve multiple functions. The same sensors that monitor vehicle operation for standard control functions are also used as input data for the shock load prediction model, eliminating the need for dedicated shock load sensors and reducing manufacturing costs.
Solution Approach 2:
Instead of installing additional physical sensors on production vehicles, the patent uses a trained digital twin model that copies the measurement capabilities of the additional sensors through virtual simulation and data processing.
3Measurement precision
If training data is collected from test vehicles with additional sensors, then relationship accuracy between CAN data and shock loads is improved, but loss of time and cost increase due to extensive testing
Solution Approach 1:
The patent performs preliminary training of the digital twin model using data from a limited set of test vehicles before deploying the system to production vehicles. This preliminary action allows the model to learn the relationships between CAN bus data and shock loads in advance, so that once trained, the system can operate without requiring further extensive testing or data collection.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The invention pertains to a method (100) for predicting failure of one or more components of an individual vehicle, the vehicle comprising a plurality of generic sensors, the method comprising: providing (110) an individual digital twin for the vehicle, the digital twin comprising virtual components that are digital representations of the one or more components of the vehicle; providing (120) digital generic sensors at the digital twin, wherein each digital generic sensor is a digital representation of one of the plurality of generic sensors and generates virtual generic data; providing (125) a digital shock load sensor at each of the virtual components, each digital shock load sensor generating virtual shock load data; performing a simulation involving the digital twin, wherein virtual generic data and virtual shock load data are generated simultaneously; monitoring (130), during the simulation, the virtual generic data and the virtual shock load data; and deriving (140) a relationship between the virtual shock load data and the virtual generic data, wherein the method further comprises: continuously monitoring (150), during a motion of the vehicle, generic sensor data generated by the plurality of generic sensors; inferring (160), based on the monitored generic sensor data and using the derived relationship, actual shock loads affecting each of the one or more components; estimating (170), based on the inferred shock loads, a damage to and/or a time to failure of the one or more components and updating (180) the individual digital twin based on the estimated damage and/or time of failure.