CN-122020348-A - Heavy truck vehicle anomaly identification analysis method based on distributed cluster and machine learning

CN122020348ACN 122020348 ACN122020348 ACN 122020348ACN-122020348-A

Abstract

The invention relates to the technical field of intelligent networking application of heavy commercial vehicles, in particular to a heavy truck vehicle anomaly identification analysis method based on distributed clusters and machine learning; the system realizes high-efficiency processing and accurate analysis of mass data by integrating an advanced distributed computing technology and a machine learning algorithm, firstly stores acquired data in a high-performance distributed file system by utilizing a distributed cluster technology, ensures high availability of the system by load balancing, secondly, pre-processes and extracts features of the data by utilizing a distributed computing framework to provide a high-quality data basis for subsequent abnormal recognition, secondly, builds an abnormal recognition model by adopting an isolated forest algorithm, realizes accurate recognition of abnormal conditions such as oil consumption, high temperature, torque limitation and the like by training and optimizing the data, and finally, displays abnormal recognition results by utilizing a visual interface and an API interface, thereby greatly improving the large data intellectualization level and the operation efficiency of the heavy truck.

Inventors

YANG ZHIGANG
ZHANG WENBO
DANG PENGCHENG
YANG JIE
LIU JIALIN

Assignees

陕西重型汽车有限公司

Dates

Publication Date: 20260512
Application Date: 20241112

Claims (6)

1. The heavy truck vehicle anomaly identification analysis method based on the distributed cluster and the machine learning is characterized by comprising a data acquisition and analysis module, a distributed storage module, a distributed calculation module, an anomaly identification module, an analysis module and an application module; The distributed storage module stores the collected mass data by using an HDFS distributed file system; The distributed computing module utilizes APACHE SPARK distributed computing frames to efficiently process and analyze data stored in a distributed file system; The anomaly identification module constructs an anomaly identification model of anomaly information through a machine learning isolated forest algorithm; the analysis module carries out deep analysis on the abnormal information and carries out joint analysis by combining the vehicle related signals.
2. The heavy truck anomaly identification analysis method based on distributed clustering and machine learning according to claim 1, wherein the distributed computing module preprocesses data, extracts statistical features and time sequence features for anomaly information identification from the original data through a feature extraction algorithm, and takes the extracted statistical features and time sequence features as input to train an isolated forest machine learning model.
3. The heavy truck anomaly identification analysis method based on distributed clustering and machine learning according to claim 1, wherein the application module is provided with a visual interface and an API interface, and can check anomaly information identification results and receive early warning information.
4. The heavy truck anomaly identification analysis method based on distributed clustering and machine learning according to claim 1, wherein the data acquisition and analysis module adopts a vehicle-mounted data terminal to transmit back and analyze vehicle operation data through a 4G network.
5. The heavy truck anomaly identification analysis method based on distributed clustering and machine learning according to claim 1, wherein the anomaly information comprises oil consumption, high temperature and torque limitation.
6. The heavy truck vehicle anomaly identification analysis method based on distributed clustering and machine learning of claim 1, wherein the vehicle related signals include engine speed, fan speed, torque, water temperature, load, throttle opening.

Description

Heavy truck vehicle anomaly identification analysis method based on distributed cluster and machine learning Technical Field The invention relates to the technical field of intelligent networking application of heavy commercial vehicles, in particular to a heavy truck vehicle anomaly identification analysis method based on distributed clusters and machine learning. Background With the rapid development of the logistics industry, heavy truck vehicles are used as main forces for logistics transportation, and the safety, economy and environmental protection of the running state of the heavy truck vehicles are important. However, the heavy truck vehicle may be affected by various factors during use, resulting in an abnormal phenomenon of the vehicle. In the field of logistics transportation, the monitoring and management of heavy truck vehicles is always an important research topic. With the development of internet of things, more and more systems begin to collect operation data of vehicles by using sensors and identify anomalies by combining data analysis technology. In the prior art, it is common practice to store and process collected operation data in a centralized manner, and then to use some statistical methods to identify anomalies. These systems, while capable of performing basic monitoring functions, suffer from the following drawbacks: 1. the data processing efficiency is low, and as the centralized data processing mode needs to transmit all data to the central server for processing, the data transmission and processing efficiency is low, and the requirement of large-scale data processing cannot be met. 2. Some systems only provide basic parameter display functions, lack active abnormality recognition, and lack in-depth analysis and processing suggestions for abnormality causes. Therefore, the method capable of actively identifying and analyzing the anomalies of the heavy truck such as fuel consumption, high temperature, torque limiting and the like is developed, and has important significance for improving the efficiency and the safety of logistics transportation. Disclosure of Invention Aiming at the problems existing in the prior art, the invention aims to provide a heavy truck anomaly identification analysis method based on distributed clustering and machine learning. The heavy truck vehicle anomaly identification analysis method based on distributed cluster and machine learning comprises a data acquisition and analysis module, a distributed storage module, a distributed calculation module, an anomaly identification module, an analysis module and an application module; The distributed storage module stores the collected mass data by using an HDFS distributed file system; The distributed computing module utilizes APACHE SPARK distributed computing frames to efficiently process and analyze data stored in a distributed file system; The anomaly identification module constructs an anomaly identification model of anomaly information through a machine learning isolated forest algorithm; the analysis module carries out deep analysis on the abnormal information and carries out joint analysis by combining the vehicle related signals. The method is characterized in that the distributed computing module preprocesses data, extracts statistical features and time sequence features for identifying abnormal information from the original data through a feature extraction algorithm, takes the extracted statistical features and time sequence features as input, and trains an isolated forest machine learning model. Preferably, the application module is provided with a visual interface and an API interface, and can check the abnormal information identification result and receive the early warning information. Preferably, the data acquisition and analysis module adopts a vehicle-mounted data terminal to transmit back and analyze vehicle operation data through a 4G network. Preferably, the abnormality information includes fuel consumption, high temperature, and torque limitation. Preferably, the vehicle related signals include engine speed, fan speed, torque, water temperature, load, and accelerator opening. The invention has the following beneficial effects: 1) Real-time performance, wherein the system can timely acquire and process operation data, timely find out anomalies such as oil consumption, high temperature, torque limitation and the like, and reduce the influence of the anomalies on operation; 2) The accuracy is that the machine learning isolated forest algorithm is adopted to carry out abnormal identification, so that the accuracy and the stability are higher; 3) Expandability, namely, large-scale data and high concurrency requests can be easily handled by using a distributed cluster architecture; 4) Intelligence, namely automatically learning and optimizing a model through a machine learning algorithm, and improving the accuracy and efficiency of anomaly identification; 5) The flexibility is that a visual inte