| CPC G07C 5/0816 (2013.01) [G06F 18/24137 (2023.01); G07C 5/0808 (2013.01)] | 8 Claims |
|
1. A vehicle diagnostic system that diagnoses a target vehicle using sound data obtained by recording a sound produced from the target vehicle, the target vehicle being a vehicle to be diagnosed, the vehicle diagnostic system comprising:
processing circuitry; and
a computer-readable medium, wherein
the computer-readable medium stores:
data of a learned model trained by supervised learning to generate operation data from training sound data using training data including the training sound data and the operation data, the training sound data being recorded while operating a reference vehicle in which no anomaly has occurred, the operation data including variables, and the operation data indicating an operation status of the reference vehicle collected simultaneously with the training sound data; and
cluster data created by outputting pieces of generated data that is the operation data generated by the learned model using pieces of sound data recorded using vehicles in which types of anomalies are different from each other and identified and by clustering, for each of the types of anomalies, a loss variable indicating a magnitude of an error in each of the variables in each piece of the generated data from each of the variables in the pieces of the operation data collected simultaneously with the sound data, wherein the loss variable is a set of sums of errors in one dataset for each of the variables included in the operation data, and
the processing circuitry is configured to execute:
a loss calculation process that inputs diagnostic sound data recorded while operating the target vehicle to the learned model, outputs the generated data, and calculates the loss variable in the generated data; and
a diagnostic process that checks data of the loss variable calculated through the loss calculation process against the cluster data, determines to which cluster the loss variable calculated through the loss calculation process belongs, and outputs a diagnosis result indicating that a type of anomaly corresponding to the cluster to which the loss variable is determined to belong has occurred in the target vehicle,
in the diagnostic process, the processing circuitry is configured to:
calculate distances between coordinates of the loss variable in a space in which coordinates are defined by a magnitude of an error in each of the variables included in the operation data and centroids of the clusters in the space;
determine that the loss variable belongs to the cluster having the centroid closest to the coordinates of the loss variable; and
that the type of anomaly corresponding to the cluster to which the loss variable is determined to belong has occurred in the target vehicle.
|