CN-121997134-A - Public safety intelligent inspection early warning method and system for unmanned aerial vehicle

CN121997134ACN 121997134 ACN121997134 ACN 121997134ACN-121997134-A

Abstract

The application relates to the technical field of unmanned aerial vehicles, in particular to an unmanned aerial vehicle public safety intelligent patrol early warning method and system. The method comprises the steps of collecting inspection data based on training parameters, then creating an early warning model, inputting pre-processing data into the early warning model, generating an inspection route of a target unmanned aerial vehicle through the early warning model, judging whether an abnormal condition exists by combining multi-source data on the inspection route, sending early warning based on the severity of the abnormal condition, finally inputting real-time data into the trained early warning model, judging whether the real-time abnormal condition exists in the real-time inspection route.

Inventors

ZHANG HUA
LU WENXUAN
ZHONG FENGLEI

Assignees

连云港市航空产业有限公司

Dates

Publication Date: 20260508
Application Date: 20260126

Claims (10)

1. The utility model provides an unmanned aerial vehicle public security intelligence inspection early warning method which is characterized in that the method includes: Setting inspection parameters of the target unmanned aerial vehicle, collecting inspection data based on the training parameters, and preprocessing the inspection data to obtain a plurality of preprocessed data; Creating an early warning model; Inputting the preprocessing data into an early warning model, generating a routing inspection route of the target unmanned aerial vehicle through the early warning model, judging whether an abnormal condition exists or not by combining multi-source data on the routing inspection route, and sending early warning based on the severity of the abnormal condition to obtain a trained early warning model; and acquiring real-time data of the real-time unmanned aerial vehicle, inputting the real-time data into the trained early warning model, and judging whether the real-time abnormal condition exists in the real-time inspection route.
2. The public safety intelligent inspection early warning method of the unmanned aerial vehicle according to claim 1, wherein the method is characterized in that the inspection parameters of the target unmanned aerial vehicle are set, inspection data are collected based on training parameters, the inspection data are preprocessed, and a plurality of preprocessed data are obtained, and the method comprises the following steps: setting an unmanned aerial vehicle database; Setting corresponding acquisition parameters for each target unmanned aerial vehicle respectively, wherein the acquisition parameters comprise a patrol area, key targets and patrol heights; Collecting inspection data of each target unmanned aerial vehicle based on the collection parameters, and inputting the collected inspection data into an unmanned aerial vehicle database, wherein the inspection data comprises image data and information data; randomly selecting a target unmanned aerial vehicle inspection parameter from an unmanned aerial vehicle inspection database; judging whether missing data exists in the inspection parameters of the target unmanned aerial vehicle; If missing data exists in the inspection parameters of the target unmanned aerial vehicle, filling the missing parameters based on the average value; And returning to randomly selecting the inspection parameters of one target unmanned aerial vehicle from the unmanned aerial vehicle inspection database until all the inspection parameters of the target unmanned aerial vehicle in the unmanned aerial vehicle database are selected, and obtaining a plurality of preprocessing data.
3. The public safety intelligent inspection early warning method of the unmanned aerial vehicle according to claim 2, wherein the preprocessing data is input into an early warning model, an inspection route of the target unmanned aerial vehicle is generated through the early warning model, whether an abnormal condition exists is judged by combining multi-source data on the inspection route, early warning is sent out based on the severity of the abnormal condition, and a trained early warning model is obtained, and the method comprises the following steps: Dividing all the preprocessed data into a training set and a testing set according to random proportion; inputting the training set into an early warning model, respectively creating a corresponding inspection route for each target unmanned aerial vehicle through the early warning model, and judging whether an abnormal condition exists on the inspection route by combining the preprocessing data; If abnormal conditions exist on the inspection route, judging the severity of the abnormal conditions, calculating the abnormality degree of the inspection route based on the abnormality degree of all key targets on the inspection route, and executing a corresponding early warning mode based on the severity to obtain a trained early warning model; and inputting the test set into the trained early warning model, and verifying whether the trained early warning model is trained.
4. The public safety intelligent inspection early warning method of the unmanned aerial vehicle according to claim 3, wherein the training set is input into an early warning model, a corresponding inspection route is respectively created for each target unmanned aerial vehicle through the early warning model, and whether an abnormal condition exists on the inspection route is judged by combining the preprocessing data, and the method comprises the following steps: randomly selecting preprocessing data of a target unmanned aerial vehicle from a training set; Acquiring a patrol route and a key target of the target unmanned aerial vehicle from the preprocessing data, and marking the key target on the patrol route; for each key target, calculating an abnormal value of each key target respectively; and returning to randomly selecting the preprocessing data of one target unmanned aerial vehicle from the training set until the preprocessing data of all the target unmanned aerial vehicles in the training set are selected, and obtaining the corresponding abnormality degree of each target unmanned aerial vehicle.
5. The public safety intelligent patrol early warning method of the unmanned aerial vehicle according to claim 4, wherein if an abnormal condition exists on a patrol route, judging the severity of the abnormal condition, calculating the abnormality of the patrol route based on the abnormality of all key targets on the patrol route, and executing a corresponding early warning mode based on the severity to obtain a trained early warning model, and the method comprises the following steps: Setting a plurality of abnormal degrees and abnormal values corresponding to the abnormal degrees; setting an abnormality degree for each key target based on the abnormality value; Calculating the abnormality degree of the inspection route based on the abnormality degrees of all key targets of each inspection route; setting an abnormality degree threshold value, and judging whether the abnormality degree of the routing inspection route is greater than or equal to the abnormality degree threshold value; And if the abnormality degree of the routing inspection route is greater than or equal to the abnormality degree threshold value, marking the routing inspection route as an abnormal route.
6. The public safety intelligent patrol early warning method of the unmanned aerial vehicle according to claim 5, wherein the collecting real-time data of the real-time unmanned aerial vehicle, inputting the real-time data into the trained early warning model, judging whether the real-time abnormal condition exists in the real-time patrol route, comprises the following steps: collecting real-time inspection data of the real-time unmanned aerial vehicle, wherein the real-time inspection data comprises real-time image data and real-time information data; inputting real-time inspection data into a trained early warning model, and calculating the abnormality degree of a real-time key target and the abnormality degree of a real-time inspection route of the real-time unmanned aerial vehicle through the trained early warning model; and judging whether the real-time routing inspection route is abnormal or not according to the abnormality degree of the real-time routing inspection route.
7. The public safety intelligent patrol early warning method of the unmanned aerial vehicle according to claim 6, wherein the step of judging whether the real-time route is abnormal or not by combining the abnormality degree of the real-time patrol route comprises the following steps: clustering analysis is carried out on the real-time routing inspection route to obtain a routing inspection route corresponding to the real-time routing inspection route; setting a route abnormality threshold; Calculating the difference value between the abnormality degree of the real-time routing inspection route and the abnormality degree of the target route, and judging whether the difference value between the abnormality degree of the real-time routing inspection route and the abnormality degree of the target route is larger than or equal to a route abnormality threshold value; if the difference value between the abnormality degree of the real-time routing inspection route and the abnormality degree of the target route is larger than or equal to the route abnormality threshold value, the real-time routing inspection route is a real-time abnormal route.
8. The unmanned aerial vehicle public safety intelligent patrol early warning system is applied to the unmanned aerial vehicle public safety intelligent patrol early warning method according to any one of claims 1 to 7, and is characterized by comprising the following steps: a target unmanned plane; The system comprises an acquisition component, a detection component and a control component, wherein the acquisition component is arranged on the target unmanned aerial vehicle and is used for respectively acquiring inspection data generated in the inspection process of the target unmanned aerial vehicle; The system comprises an acquisition component, an early warning component, wherein the early warning component is in communication connection with the acquisition component, all data information acquired by the acquisition component is input to the early warning component, whether an abnormal condition exists is judged by the early warning component through combining multi-source data on a routing inspection route, and early warning is sent out based on the severity of the abnormal condition.
9. The unmanned aerial vehicle public safety intelligent patrol early-warning system according to claim 8, wherein the early-warning assembly comprises a processor and a storage unit, patrol data is stored by the storage unit, and the unmanned aerial vehicle public safety intelligent patrol early-warning method is executed by the processor.
10. The unmanned aerial vehicle public safety intelligent patrol early warning system according to claim 9, wherein the acquisition assembly comprises an image acquisition module and an information acquisition module, wherein the image acquisition module is used for acquiring patrol images of patrol routes, and the information acquisition module is used for acquiring patrol information of the patrol routes.

Description

Public safety intelligent inspection early warning method and system for unmanned aerial vehicle Technical Field The invention relates to the technical field of unmanned aerial vehicles, in particular to an unmanned aerial vehicle public safety intelligent patrol early warning method and system. Background The unmanned aerial vehicle inspection is a technology ‌ for carrying various sensors on an unmanned aerial vehicle to automatically inspect a target area or facility and acquire and analyze data, and the unmanned aerial vehicle inspection system is capable of improving efficiency, guaranteeing safety and realizing intellectualization, and deeply changes the traditional inspection modes in various fields such as power grid, traffic, environmental protection, urban management and the like. The invention discloses a Chinese patent with publication number of CN118692166A, which discloses an unmanned aerial vehicle inspection system and an unmanned aerial vehicle inspection method, and relates to the technical field of unmanned aerial vehicle inspection, the system comprises a tower region dividing module, a monitoring data determining module, an inspection equipment screening module, an inspection equipment analyzing module, an unmanned aerial vehicle screening module, a tower state analyzing module and an early warning terminal, each normal region and each key region are obtained through dividing, inspection equipment corresponding to each normal region and each key region is obtained through screening, and the inspection duration and the inspection equipment energy consumption corresponding to each normal area and each key area are predicted, so that the corresponding inspection unmanned aerial vehicle is obtained through screening, and after the unmanned aerial vehicle inspection is finished, the corresponding state of the power tower is judged, so that the safety of the power tower is ensured, but in the prior art, judgment is not made according to key targets contained in the inspection route, and the judgment on the abnormal condition of the user inspection route is easy to generate errors. Disclosure of Invention The invention aims to provide an unmanned aerial vehicle public safety intelligent patrol early warning method and system, which are used for solving the problems in the background technology. In order to solve the technical problems, one of the purposes of the present invention is to provide an unmanned aerial vehicle public security intelligent patrol and early warning method, comprising: Setting inspection parameters of the target unmanned aerial vehicle, collecting inspection data based on the training parameters, and preprocessing the inspection data to obtain a plurality of preprocessed data; Creating an early warning model; Inputting the preprocessing data into an early warning model, generating a routing inspection route of the target unmanned aerial vehicle through the early warning model, judging whether an abnormal condition exists or not by combining multi-source data on the routing inspection route, and sending early warning based on the severity of the abnormal condition to obtain a trained early warning model; and acquiring real-time data of the real-time unmanned aerial vehicle, inputting the real-time data into the trained early warning model, and judging whether the real-time abnormal condition exists in the real-time inspection route. Preferably, the setting of the inspection parameters of the target unmanned aerial vehicle acquires inspection data based on the training parameters, and performs preprocessing on the inspection data to obtain a plurality of preprocessed data, including the following steps: setting an unmanned aerial vehicle database; Setting corresponding acquisition parameters for each target unmanned aerial vehicle respectively, wherein the acquisition parameters comprise a patrol area, key targets and patrol heights; Collecting inspection data of each target unmanned aerial vehicle based on the collection parameters, and inputting the collected inspection data into an unmanned aerial vehicle database, wherein the inspection data comprises image data and information data; randomly selecting a target unmanned aerial vehicle inspection parameter from an unmanned aerial vehicle inspection database; judging whether missing data exists in the inspection parameters of the target unmanned aerial vehicle; If missing data exists in the inspection parameters of the target unmanned aerial vehicle, filling the missing parameters based on the average value; And returning to randomly selecting the inspection parameters of one target unmanned aerial vehicle from the unmanned aerial vehicle inspection database until all the inspection parameters of the target unmanned aerial vehicle in the unmanned aerial vehicle database are selected, and obtaining a plurality of preprocessing data. Preferably, the preprocessing data is input to an early warning model, a routing inspectio