CN-122009091-A - Safe car washing method, device, equipment and medium

CN122009091ACN 122009091 ACN122009091 ACN 122009091ACN-122009091-A

Abstract

The invention discloses a safe vehicle washing method, a device, equipment and a medium, which relate to the technical field of vehicle washing and comprise the steps of combining control logic of vehicle washing according to a vehicle washing control instruction to acquire vehicle washing data; and calling a fault early warning algorithm model to execute vehicle safety cleaning operation according to the abnormal vehicle cleaning recognition result to obtain a safety vehicle cleaning result. The method effectively improves the accuracy of abnormality identification and the timeliness of risk treatment in the automatic car washing process, reduces the occurrence probability of potential safety hazards such as equipment faults, car scratch and the like, ensures the continuity, stability and intelligence level of the car washing process, and simultaneously accumulates effective data support for iterative optimization of an algorithm model.

Inventors

LUO GONGBO
LI CHANGFU
Yan Zhiya
Cui Longkuan

Assignees

三盈联合科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260121

Claims (10)

1. A method of safely washing a vehicle, comprising: s1, acquiring vehicle cleaning data according to a vehicle cleaning control instruction and a control logic of vehicle cleaning; s2, carrying out abnormality recognition by utilizing a multi-mode fusion recognition model based on the vehicle cleaning data to obtain an abnormality recognition result of vehicle cleaning; and S3, calling a fault early warning algorithm model to execute the safe vehicle cleaning operation according to the abnormal recognition result of the vehicle cleaning, and obtaining a safe vehicle cleaning result.
2. The method of claim 1, wherein S1, according to the vehicle washing control command, in combination with the control logic of the vehicle washing, obtains vehicle washing data, comprising: Generating a vehicle cleaning instruction by combining the vehicle identification result with the vehicle cleaning trigger signal, and loading vehicle cleaning control parameters; Based on the vehicle cleaning instruction and the vehicle cleaning control parameter, triggering control logic of vehicle cleaning, and collecting vehicle cleaning image data and vehicle cleaning running state data; Performing association matching on the vehicle cleaning image data and the vehicle cleaning running state data, and preprocessing to obtain preprocessed multi-dimensional vehicle cleaning data; And carrying out structural processing on the preprocessed multidimensional vehicle cleaning data to obtain the vehicle cleaning data.
3. The safe vehicle washing method according to claim 1, wherein S2, based on the vehicle washing data, performs abnormality recognition by using a multi-mode fusion recognition model, and obtains an abnormality recognition result of vehicle washing, comprising: Performing initial visual abnormality identification on the vehicle cleaning data by adopting an edge algorithm to obtain initial visual abnormality data of the vehicle cleaning data; performing initial equipment abnormality identification according to the vehicle cleaning data and a preset vehicle cleaning parameter threshold value, and acquiring initial equipment abnormality data of the vehicle cleaning data; and acquiring an abnormal recognition result of vehicle cleaning by utilizing a multi-mode fusion recognition model based on the initial visual abnormal data of the vehicle cleaning data and the initial equipment abnormal data of the vehicle cleaning data.
4. A safe vehicle washing method according to claim 3, wherein obtaining an abnormality recognition result of vehicle washing based on initial visual abnormality data of the vehicle washing data and initial equipment abnormality data of the vehicle washing data using a multi-modal fusion recognition model, comprises: Extracting features of the initial visual abnormality data of the vehicle cleaning data and the initial equipment abnormality data of the vehicle cleaning data by utilizing a multi-mode fusion recognition model to obtain initial visual abnormality features and initial equipment abnormality features of the vehicle cleaning data; performing cross-modal association fusion by adopting a weighted fusion algorithm based on the initial visual abnormal characteristics and the initial equipment abnormal characteristics of the vehicle cleaning data to obtain a fusion characteristic vector of the vehicle cleaning data; obtaining a classification abnormality recognition result through a pre-trained classification rule according to the fusion feature vector of the vehicle cleaning data; And matching the classified abnormal recognition result with a preset normal cleaning database to obtain an abnormal recognition result of vehicle cleaning.
5. The method for safely washing vehicles according to claim 3, wherein S3, calling a fault early warning algorithm model to execute a safe vehicle washing operation according to the abnormal recognition result of the vehicle washing, and obtaining a safe vehicle washing result comprises: Performing risk level analysis by using a fault early warning algorithm model based on the abnormal recognition result of the vehicle cleaning to obtain a risk level analysis result of the vehicle cleaning; Generating a safety control instruction according to the dangerous grade analysis result of the vehicle cleaning; and executing the safe vehicle cleaning operation by utilizing the safe control instruction to obtain a safe vehicle cleaning result.
6. The method of claim 5, wherein the step of performing a risk level analysis based on the abnormality recognition result of the vehicle cleaning using a failure early warning algorithm model to obtain a risk level analysis result of the vehicle cleaning comprises: Inputting the abnormal recognition result of the vehicle cleaning into the fault early warning algorithm model for structural disassembly to obtain standardized abnormal element data; Performing primary judgment on the standardized abnormal element data by using a preset risk classification standard to acquire a risk level of vehicle cleaning; Judging whether the risk level of the vehicle cleaning is a fault level, if so, acquiring the fault level vehicle cleaning as a risk level analysis result of the vehicle cleaning, otherwise, executing a first operation; The first operation is to carry out secondary judgment on the standardized abnormal element data based on a non-fault risk classification standard, and obtain a non-fault risk level of vehicle cleaning as a risk level analysis result of vehicle cleaning.
7. The method of claim 6, wherein performing a vehicle safety washing operation using the safety control command to obtain a safety washing result comprises: acquiring an adjustment strategy for vehicle cleaning according to the safety control instruction; Based on the safety control instruction and the adjustment strategy of the vehicle cleaning, executing the vehicle safety cleaning operation by combining the control logic of the vehicle cleaning, and acquiring a vehicle cleaning state; Judging whether the vehicle cleaning state is a fault state, if so, acquiring the risk level of vehicle fault cleaning according to the vehicle cleaning state, and executing a second operation, otherwise, executing a third operation; Judging whether the risk level of the vehicle fault cleaning is a fault level, if so, generating an emergency stop instruction, combining the vehicle cleaning state to obtain a safe vehicle cleaning result, otherwise, re-acquiring vehicle cleaning data, and executing a fourth operation; Judging whether the vehicle cleaning state is a cleaning completion state, if so, acquiring a safe vehicle cleaning result by using the vehicle cleaning state, otherwise, executing a vehicle safe cleaning operation according to the vehicle cleaning state in combination with the control logic of the vehicle cleaning, collecting vehicle cleaning data in real time, and executing the fourth operation; and the fourth operation is to perform initial visual abnormality identification on the vehicle cleaning data by adopting an edge algorithm, and obtain initial visual abnormality data of the vehicle cleaning data.
8. A safe car washing device, which is characterized by comprising a car washing machine core controller and a distributed AI intelligent analysis terminal, wherein the car washing machine core controller comprises a data acquisition module and a core control module; the data acquisition module is used for acquiring vehicle cleaning data according to a vehicle cleaning control instruction and a vehicle cleaning control logic; the distributed AI intelligent analysis terminal is used for carrying out abnormality identification by utilizing a multi-mode fusion identification model based on the vehicle cleaning data to obtain an abnormality identification result of vehicle cleaning; and the core control module is used for calling a fault early warning algorithm model to execute the vehicle safety cleaning operation according to the abnormal recognition result of the vehicle cleaning, and obtaining a safety vehicle cleaning result.
9. An electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-7.
10. A computer readable storage medium, having stored thereon a computer program, wherein the computer program when executed by one or more processors implements the method of any of claims 1-7.

Description

Safe car washing method, device, equipment and medium Technical Field The invention relates to the technical field of vehicle cleaning, in particular to a safe vehicle cleaning method, device, equipment and medium. Background Traditional car washing equipment relies on manual operation, manual inspection or single sensor monitoring more, and not only lacks the comprehensive perceptibility of car washing scenes, but also is difficult to realize multi-dimensional real-time control of equipment states. For equipment operation states such as abrupt change of high-voltage motor current and abnormal pressure of a cylinder, the monitoring dimension of a single sensor is limited, potential faults cannot be early warned, the probability of safety accidents such as personnel accidental injury and equipment damage in the car washing process is high, the fault response is delayed, the safety, stability and continuity of car washing operation are difficult to guarantee, and the requirements of the modern intelligent car washing automation and high safety standard cannot be met. Therefore, a need exists for a safe car washing method, apparatus, device and medium that overcomes the shortcomings of the prior art. Disclosure of Invention The invention aims to provide a safe car washing method, device, equipment and medium, which are used for solving the problems that the traditional car washing equipment is easy to cause personnel injury or equipment damage in the car washing process due to scene anomaly identification hysteresis and equipment failure early warning deficiency, realizing full-dimension perception, risk analysis and dynamic response of the car washing scene and equipment state, and improving the safety and intelligent level of the car washing process. In a first aspect, to achieve the above object, the present invention provides a safe car washing method, including: s1, acquiring vehicle cleaning data according to a vehicle cleaning control instruction and a control logic of vehicle cleaning; s2, carrying out abnormality recognition by utilizing a multi-mode fusion recognition model based on the vehicle cleaning data to obtain an abnormality recognition result of vehicle cleaning; and S3, calling a fault early warning algorithm model to execute the safe vehicle cleaning operation according to the abnormal recognition result of the vehicle cleaning, and obtaining a safe vehicle cleaning result. Optionally, S1, acquiring vehicle cleaning data according to a vehicle cleaning control instruction in combination with a vehicle cleaning control logic, including: Generating a vehicle cleaning instruction by combining the vehicle identification result with the vehicle cleaning trigger signal, and loading vehicle cleaning control parameters; Based on the vehicle cleaning instruction and the vehicle cleaning control parameter, triggering control logic of vehicle cleaning, and collecting vehicle cleaning image data and vehicle cleaning running state data; Performing association matching on the vehicle cleaning image data and the vehicle cleaning running state data, and preprocessing to obtain preprocessed multi-dimensional vehicle cleaning data; And carrying out structural processing on the preprocessed multidimensional vehicle cleaning data to obtain the vehicle cleaning data. Optionally, S2, performing anomaly identification by using a multi-mode fusion identification model based on the vehicle cleaning data, to obtain an anomaly identification result of vehicle cleaning, including: Performing initial visual abnormality identification on the vehicle cleaning data by adopting an edge algorithm to obtain initial visual abnormality data of the vehicle cleaning data; performing initial equipment abnormality identification according to the vehicle cleaning data and a preset vehicle cleaning parameter threshold value, and acquiring initial equipment abnormality data of the vehicle cleaning data; and acquiring an abnormal recognition result of vehicle cleaning by utilizing a multi-mode fusion recognition model based on the initial visual abnormal data of the vehicle cleaning data and the initial equipment abnormal data of the vehicle cleaning data. Optionally, based on the initial visual anomaly data of the vehicle cleaning data and the initial equipment anomaly data of the vehicle cleaning data, a multi-mode fusion recognition model is used to obtain an anomaly recognition result of vehicle cleaning, including: Extracting features of the initial visual abnormality data of the vehicle cleaning data and the initial equipment abnormality data of the vehicle cleaning data by utilizing a multi-mode fusion recognition model to obtain initial visual abnormality features and initial equipment abnormality features of the vehicle cleaning data; performing cross-modal association fusion by adopting a weighted fusion algorithm based on the initial visual abnormal characteristics and the initial equipment abnormal characteristics of the