CN114863688A

CN114863688A - Intelligent positioning method and system for muck vehicle based on unmanned aerial vehicle

Info

Publication number: CN114863688A
Application number: CN202210786753.4A
Authority: CN
Inventors: 杨翰翔; 肜卿
Original assignee: Shenzhen Lianhe Intelligent Technology Co ltd
Current assignee: Shenzhen Lianhe Intelligent Technology Co ltd
Priority date: 2022-07-06
Filing date: 2022-07-06
Publication date: 2022-08-05
Anticipated expiration: 2042-07-06
Also published as: CN114863688B

Abstract

The invention provides an intelligent muck vehicle positioning method and system based on an unmanned aerial vehicle, which can improve the reliability of the collection of the travelling path information by acquiring a monitoring data stream for monitoring the muck vehicle in a target control area by the monitoring unmanned aerial vehicle and controlling a plurality of unmanned aerial vehicles to collect the travelling path information by linkage when the monitoring data stream is detected to indicate that the target muck vehicle has violation behaviors, can carry out travelling positioning prediction on the travelling path information collected by the unmanned aerial vehicles, can conveniently and accurately analyze the travelling positioning prediction information of the target muck vehicle in a later preset time period, and sends the travelling positioning prediction information to a patrol police duty service terminal corresponding to the target control area for early warning information prompt, thereby improving the early warning timeliness and accuracy of a construction site.

Description

Intelligent positioning method and system for muck vehicle based on unmanned aerial vehicle

Technical Field

The invention relates to the technical field of unmanned aerial vehicle monitoring, in particular to an intelligent positioning method and system for a muck vehicle based on an unmanned aerial vehicle.

Background

For the construction site on site, the working environment is complex and changeable, safety accidents are often easy to occur, and how to implement real-time comprehensive management of the construction site is a very troublesome problem for relevant units. For example, regarding the violation behaviors of the soil inlet and outlet vehicles, if the related violation behaviors are not discovered in time and early warning is given, a major safety accident may be caused. However, in the current monitoring scheme, the early warning mode of simple area comparison is performed only after the single unmanned aerial vehicle acquires the driving path information, and the early warning effect is poor.

Disclosure of Invention

In order to overcome at least the defects in the prior art, the invention aims to provide an intelligent positioning method and system for a muck vehicle based on an unmanned aerial vehicle.

In a first aspect, the invention provides an intelligent muck vehicle positioning method based on an unmanned aerial vehicle, which is applied to a cloud service platform, wherein the cloud service platform is in communication connection with a plurality of monitoring unmanned aerial vehicles, and the method comprises the following steps:

acquiring a monitoring data stream for monitoring the muck vehicle in a target control area by the monitoring unmanned aerial vehicle;

when the monitoring data stream is detected to indicate that the target muck truck has violation behaviors, other unmanned aerial vehicles in the target control area are controlled in a linkage mode to acquire driving path information of the target muck truck in real time;

and predicting driving positioning prediction information of the target slag car in a later preset time period based on the driving path information of the target slag car collected by each unmanned aerial vehicle, and sending the driving positioning prediction information to a patrol duty service terminal corresponding to the target management and control area for early warning information prompt.

In a second aspect, an embodiment of the present invention further provides an intelligent positioning system for a muck vehicle based on an unmanned aerial vehicle, where the intelligent positioning system for a muck vehicle based on an unmanned aerial vehicle comprises a cloud service platform and a plurality of monitoring unmanned aerial vehicles in communication connection with the cloud service platform;

the cloud service platform is used for:

According to any one of the aspects, in the embodiment provided by the invention, the monitoring data stream for monitoring the muck truck in the target control area by the monitoring unmanned aerial vehicle is obtained, when the monitoring data stream is detected to indicate that the target muck truck has violation behaviors, the plurality of unmanned aerial vehicles are controlled in a linkage manner to acquire the driving path information, so that the reliability of the driving path information acquisition can be improved, the driving positioning prediction can be performed according to the driving path information acquired by the plurality of unmanned aerial vehicles, the driving positioning prediction information of the target muck truck in a later preset time period can be conveniently and accurately analyzed, and the driving positioning prediction information is sent to the patrol duty service terminal corresponding to the target control area for early warning information prompt, so that the early warning timeliness and accuracy of a construction site are improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that need to be called in the embodiments are briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention, and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.

Fig. 1 is a schematic view of an application scenario of an intelligent positioning system of a muck vehicle based on an unmanned aerial vehicle according to an embodiment of the present invention.

Fig. 2 is a schematic flow chart of the intelligent positioning method for the muck vehicle based on the unmanned aerial vehicle according to the embodiment of the invention.

Fig. 3 is a block diagram schematically illustrating a structure of a cloud service platform for implementing the above-described intelligent positioning method for a muck vehicle based on an unmanned aerial vehicle according to an embodiment of the present invention.

Detailed Description

The present invention is described in detail below with reference to the drawings, and the specific operation methods in the method embodiments can also be applied to the apparatus embodiments or the system embodiments.

Fig. 1 is a schematic view of an application scenario of an intelligent positioning system 10 for a muck truck based on an unmanned aerial vehicle according to an embodiment of the present invention. The unmanned aerial vehicle-based muck vehicle intelligent positioning system 10 can include a cloud service platform 100 and a monitoring unmanned aerial vehicle 200 in communication with the cloud service platform 100. The intelligent drone-based muck vehicle positioning system 10 shown in fig. 1 is merely one possible example, and in other possible embodiments, the intelligent drone-based muck vehicle positioning system 10 may also include only at least some of the components shown in fig. 1 or may also include other components.

In a possible design idea, the cloud service platform 100 and the monitoring drone 200 in the unmanned aerial vehicle-based muck vehicle intelligent positioning system 10 may execute the unmanned aerial vehicle-based muck vehicle intelligent positioning method described in the following method embodiment in a matching manner, and the execution steps of the cloud service platform 100 and the monitoring drone 200 may be described in detail in the following method embodiment.

In order to solve the technical problem in the foregoing background art, the method for intelligently positioning a muck vehicle based on an unmanned aerial vehicle provided in this embodiment may be executed by the cloud service platform 100 shown in fig. 1, and the method for intelligently positioning a muck vehicle based on an unmanned aerial vehicle is described in detail below. Fig. 2 is a schematic flow chart of an intelligent positioning method for a muck vehicle based on an unmanned aerial vehicle according to an embodiment of the present invention, where the intelligent positioning method for a muck vehicle based on an unmanned aerial vehicle includes:

step S110, acquiring a monitoring data stream for monitoring the muck vehicle in the target control area by the monitoring unmanned aerial vehicle.

And step S120, when the fact that the monitoring data stream indicates that the target muck truck has violation behaviors is detected, other unmanned aerial vehicles in the target control area are controlled in a linkage mode to acquire driving path information of the target muck truck in real time.

Step S130, predicting driving positioning prediction information of the target muck truck in a later preset time period based on the driving path information of the target muck truck acquired by each unmanned aerial vehicle, and sending the driving positioning prediction information to a patrol duty service terminal corresponding to the target management and control area for early warning information prompt.

Based on the above steps, in this embodiment, the monitoring data stream for monitoring the muck truck in the target control area by the monitoring unmanned aerial vehicle is acquired, when it is detected that the monitoring data stream indicates that the target muck truck has a violation behavior, the multiple unmanned aerial vehicles are controlled to perform driving path information acquisition in a linkage manner, reliability of driving path information acquisition can be improved, driving positioning prediction is performed on the driving path information acquired by the multiple unmanned aerial vehicles, driving positioning prediction information in a later preset time period of the target muck truck can be conveniently and accurately analyzed, and the driving positioning prediction information is sent to the patrol police service terminal corresponding to the target control area for warning information prompt, so that warning timeliness and accuracy of a construction site are improved.

In a possible design idea, for the specific implementation manner of the step S120, in which other unmanned aerial vehicles in the target control area acquire the driving path information of the target muck truck in real time through linkage control, the following steps can be implemented.

Step S121, acquiring flight state information of all candidate unmanned aerial vehicles in the target control area;

step S122, screening all candidate unmanned aerial vehicles according to the flight state information to obtain target unmanned aerial vehicles, and generating corresponding linkage control strategy information for each target unmanned aerial vehicle according to the current driving direction information of the target muck vehicle and the flight state information of each target unmanned aerial vehicle;

and S123, controlling each corresponding target unmanned aerial vehicle to acquire the driving path information of the target muck truck in real time based on each corresponding linkage control strategy information.

In a possible design idea, a specific implementation manner of predicting the driving location prediction information of the target slag car within a later preset time period based on the driving path information of the target slag car acquired by each unmanned aerial vehicle in step S130 may be implemented through the following steps.

Step S131, extracting a driving path characteristic sequence of the driving path information of the target muck truck acquired by each unmanned aerial vehicle, and fusing the driving path characteristic sequence of each driving path information with a corresponding flight state characteristic sequence of the unmanned aerial vehicle respectively to obtain a fused path characteristic cluster formed by each fused path characteristic sequence;

step S132, inputting the fusion path feature cluster into a driving positioning prediction network obtained by pre-training, and predicting driving positioning prediction information of the target muck vehicle in a later preset time period;

in a possible embodiment, for step S132, the embodiment further provides a training method based on an artificial intelligence driving location prediction network, which may include the following steps.

Step S101, obtaining a sample fusion path feature cluster and sample driving positioning prediction information corresponding to the sample fusion path feature cluster, and performing path feature description splitting on the sample fusion path feature cluster to obtain a path feature description set corresponding to driving path track information; the path feature description set corresponding to the driving path track information comprises a plurality of path feature descriptions.

In a possible design idea, splitting the path feature description of the sample fusion path feature cluster to obtain a path feature description set corresponding to the driving path trajectory information includes: and for each group of sample fusion path feature clusters in the traffic path track information, performing path feature description splitting on the traffic path track information of the sample fusion path feature clusters according to corresponding path track catastrophe points to obtain a path feature description set corresponding to the traffic path track information of the sample fusion path feature clusters. For example, the path trajectory mutation point may be set according to different mutation matching rules.

Furthermore, the step of splitting the path feature description of the vehicle path trajectory information of the sample fusion path feature cluster according to the corresponding path trajectory mutation point to obtain a path feature description set corresponding to the vehicle path trajectory information of the sample fusion path feature cluster may include the following steps: extracting path mutation characteristics of dynamic path track mutation points and static path track mutation points from the path mutation area information of the sample fusion path characteristic cluster, and generating a plurality of driving path state transition characteristics based on the extracted path mutation characteristics; according to the state transition labels of each group of the driving path state transition characteristics, determining a target decomposition rule of the state transition labels based on a track tracking decomposition rule of corresponding path track break points; and decomposing the traffic path track information of the sample fusion path feature cluster according to a target decomposition rule determined by the state transition labels corresponding to each group of traffic path state transition features to obtain a path feature description set corresponding to the traffic path track information of the sample fusion path feature cluster. In a possible design idea, the path abrupt change feature can distinguish path track abrupt change points of different segments, the driving path state transition feature is used to characterize transition state information between different driving path segments, and the target decomposition rule may include different decomposition indication information, such as according to what manner or what criteria to decompose.

In a possible design idea, before determining a target decomposition rule of the state transition label based on a trajectory tracking decomposition rule of a corresponding path trajectory mutation point according to the state transition label of each set of the driving path state transition features, the method further includes a preprocessing step of preprocessing each set of the driving path state transition features in the plurality of driving path state transition features and the corresponding state transition label. On the basis, the pretreatment step comprises the following steps: determining the state transition times of the state transition labels of each group of driving path state transition characteristics and the statistic value of the same state transition label of the same driving path state transition characteristics; cleaning the traffic route state transition characteristics of which the state transition times of the state transition labels exceed a preset transition time threshold and the statistical value of the same state transition labels exceed a preset statistical threshold corresponding to the preset transition time threshold so as to obtain the remaining traffic route track information after cleaning; filtering noise state transition labels corresponding to the rest of the cleaned path track information, and updating state transition labels with abnormal state transition in the rest of the cleaned path track information; and obtaining the driving path state transfer characteristics for path characteristic description splitting based on the rest driving path track information after cleaning and the updated state transfer label. By the design, the driving path state transition characteristics and the timeliness of the corresponding state transition labels can be ensured by pre-preprocessing, and a more accurate training data set is provided conveniently.

In another possible design idea, before performing path feature description splitting on the vehicle path trajectory information of the sample fusion path feature cluster according to the corresponding path trajectory mutation point for each group of sample fusion path feature clusters in the vehicle path trajectory information to obtain a path feature description set corresponding to the vehicle path trajectory information of the sample fusion path feature cluster, the scheme may further include the following contents: acquiring path mutation region information of a plurality of previous fusion path feature information; according to prediction feedback information of sample driving positioning prediction information, carrying out path mutation quality analysis on the path mutation area information of the plurality of previous fusion path characteristic information to obtain evaluation information of each group of previous fusion path characteristic information, wherein the evaluation information is used for screening the sample fusion path characteristic cluster from the previous fusion path characteristic information; and screening partial previous fusion path characteristic information as a sample fusion path characteristic cluster according to the evaluation information of the plurality of previous fusion path characteristic information. By the design, evaluation information can be considered when the sample fusion path feature cluster is determined, so that the quality of the fusion path feature can be considered when information prediction is carried out during subsequent network training.

In one possible design approach, the screening of the partial previous fusion path feature information as the sample fusion path feature cluster based on the evaluation information of the plurality of previous fusion path feature information includes at least one of the following embodiments.

In a first embodiment, among the plurality of previous fused path feature information, the fused path feature information before the part where the evaluation metric value corresponding to the evaluation information is higher than the preset evaluation metric value is screened as the sample fused path feature cluster.

In the second embodiment, the plurality of previous fusion path feature information are sorted in descending order according to the evaluation metric value corresponding to the evaluation information of the plurality of previous fusion path feature information, and the fusion path feature information before the part which is sorted in the front and the number of which is the set number is screened as the sample fusion path feature cluster.

And step S102, carrying out path mutation node analysis on a path feature description set corresponding to the driving path track information to obtain different path feature descriptions, and training a driving location prediction network through an input data sequence formed by the different path feature descriptions and the sample driving location prediction information. For example, the different path profiles may include representational path profiles describing the merged path feature cluster with a greater degree of significance of the corresponding path feature, non-representational path profiles describing the merged path feature cluster with a smaller degree of significance of the corresponding path feature, and/or reference path profiles, and the driving location prediction network may be a deep learning network.

In a possible design approach, in order to ensure the prediction accuracy of the driving location prediction network after training, the different path feature descriptions need to be accurately divided, and to achieve this, the content described in step S102 may be implemented by following steps S1021 to S1025.

And step S1021, determining the path characteristic description of which the path azimuth change times exceed the target time threshold as the path characteristic description set corresponding to the to-be-determined path track information for a plurality of path characteristic descriptions included in the path characteristic description set corresponding to the path track information of each group of sample fusion path characteristic clusters. In a possible design idea, the number of times of the change of the path azimuth can be used for representing the number of times of the change of the path feature description, and the more the number of times, the more important the driving path trajectory information corresponding to the path feature description is, so that the undetermined driving path trajectory information can also be understood as the driving path trajectory information focused on in the training process.

Further, this step may include the following: screening the path feature description for multiple times in path feature description sets corresponding to multiple driving path track information of different sample fusion path feature clusters, and gathering and splicing the multiple path feature descriptions screened each time to obtain multiple different path feature description sets; the method comprises the steps that path feature descriptions in a path feature description set corresponding to driving path track information of different sample fusion path feature clusters are covered in a plurality of path feature descriptions screened each time; and screening the path feature description set with the path azimuth change times exceeding the target time threshold value in a path feature description group formed by the different path feature description sets to serve as a path feature description set corresponding to the path track information of the undetermined vehicle. It is understood that the target number threshold may be adjusted according to the actual application requirements.

Step S1022, performing key feature matching processing on the path feature description in the path feature description set corresponding to the to-be-determined route track information, so as to use the route feature description meeting the key feature matching condition as an expressive route feature description representing the route track floating change of the route track information. In one possible design approach, key features may be used to differentiate between different path profiles. The route floating change of the traffic route information is used for representing the changed route information of the traffic route information when the floating change occurs.

Further, the performing key feature matching processing on the path feature description in the path feature description set corresponding to the information of the track of the vehicle to be determined may include: and extracting a plurality of path feature descriptions from the path feature description set corresponding to the path track information of the undetermined vehicle. Based on this, whether the path feature description satisfies the key feature matching condition may be implemented as follows.

Firstly, the extracted multiple path characteristics are described to execute at least one of the following processing modes:

(1) performing key feature matching on local feature descriptions in the path feature descriptions to obtain a key feature matching result aiming at the local feature descriptions;

(2) performing key feature matching on an extended feature vector set of feature vectors corresponding to each group of path feature descriptions in the plurality of path feature descriptions to obtain a key feature matching result for the extended feature vector set;

(3) respectively adding time-space domain feature marks in the path feature descriptions to perform key feature matching to obtain key feature matching results aiming at the time-space domain features;

(4) and performing key feature matching on the path feature descriptions according to the shared feature description part among the path feature descriptions to obtain a key feature matching result aiming at the shared feature description.

Secondly, if the key confidence corresponding to at least one key feature matching result is greater than the preset confidence, determining the path feature description corresponding to the at least one key feature matching result as the path feature description meeting the key feature matching condition. In the scheme, the numerical range of the key confidence coefficient can be 0-1.

By the design, whether the corresponding path feature description meets the key feature matching condition can be judged based on the key confidence of the key feature matching results of different layers, so that the omission of the expressive path feature description can be avoided.

In step S1023, a plurality of non-representational path profiles are determined, where the non-representational path profiles are path profiles included in path profiles whose path orientation change times are not greater than the target time threshold value.

Step S1024, screening partial non-expressive path characteristic descriptions as reference path characteristic descriptions according to the related characteristic descriptions among the non-expressive path characteristic descriptions in the plurality of non-expressive path characteristic descriptions. For example, the relevant profiles may be a coverage of relevant features between different non-representational path profiles, and reference to a path profile may be understood as a non-representational path profile that may be a representational path profile, i.e. a path profile that is located between a representational path profile and a non-representational path profile.

Further, this step may include the following: determining the validity times of the plurality of non-expressive path characteristic descriptions, and cleaning the non-expressive path characteristic descriptions with the validity times lower than the preset times; collecting and splicing the rest non-expressive path characteristic descriptions after cleaning to obtain a non-expressive path characteristic description set; determining a correlation profile between two random sets of non-representational path profiles in the set of non-representational path profiles; determining a local correlation feature description of each non-representational path feature description in the non-representational path feature description set and the non-representational path feature description set according to the correlation feature description between the random two sets of non-representational path feature descriptions; and according to the local relevant feature descriptions, descending the non-expressive path feature descriptions in the non-expressive path feature description set according to the content relevance corresponding to the local relevant feature descriptions, and screening the part of the non-expressive path feature descriptions which are ranked in the front as reference path feature descriptions.

Step S1025, training a driving location prediction network based on the expressive path feature description, the reference path feature description and the sample driving location prediction information of each group of sample fusion path feature clusters in the driving path track information.

In a possible design concept, the driving location prediction network includes a plurality of driving location analysis sub-networks, and further, the method for training the driving location prediction network may include the following steps: combining the expressive path characteristic description, the reference path characteristic description and the sample driving positioning prediction information of each group of sample fusion path characteristic clusters in the driving path track information into a training data sequence; training the plurality of driving location analysis sub-networks based on the composed plurality of training data sequences; and cascading the trained plurality of driving positioning analysis sub-networks through the driving positioning analysis dimensionality to obtain the driving positioning prediction network.

And S103, carrying out driving positioning prediction based on the trained driving positioning prediction network.

For example, in a possible design concept, in the process of generating corresponding linkage control strategy information for each target unmanned aerial vehicle according to the current driving direction information of the target muck vehicle and the flight state information of each target unmanned aerial vehicle, the following exemplary steps can be implemented.

Step S201, obtaining the current driving approaching direction information of each target unmanned aerial vehicle according to the current driving direction information of the target muck vehicle and the flight state information of each target unmanned aerial vehicle.

Step S202, according to the current driving approaching position information and the driving path position information of the target muck truck, acquiring a plurality of linkage control routes with preset route state labels from a linkage control route library of the target control area as undetermined linkage control routes.

Step S203, clustering the undetermined linkage control routes according to the direction relation information between the route approach direction information corresponding to each undetermined linkage control route and the current driving approach direction information of the target unmanned aerial vehicle to obtain an undetermined linkage control route cluster.

And step S204, sequentially walking each undetermined linkage control route from the first undetermined linkage control route of the undetermined linkage control route cluster according to the arrangement sequence of each undetermined linkage control route in the undetermined linkage control route cluster.

Step S205, aiming at the undetermined linkage control route which walks each time, the traveling characteristic component network corresponding to the undetermined linkage control route is respectively matched with each traveling characteristic component network corresponding to the flyable partition of the target driving path partition, and the undetermined linkage control route matched with the traveling characteristic component network corresponding to any flyable partition is used as an available linkage control route.

And step S206, judging whether the number of the currently determined available linkage control routes reaches the target number, finishing the walking operation of the pending linkage control routes when the number reaches the target number, and continuing walking the next pending linkage control route if the number does not reach the target number.

Wherein the target number is obtained by:

the method comprises the steps of obtaining reference linkage control flight data for linkage control flight of different target unmanned aerial vehicles in a preset tracking stage, wherein the reference linkage control flight data comprise recorded data, corresponding to target driving path partitions, of current planned linkage control routes used by linkage control routes planned by each target unmanned aerial vehicle in different linkage control flight processes at each time, and the recorded data comprise the number of the planned linkage control routes used in the linkage control flight processes at each time.

And taking the maximum number of used planning linkage control routes in the linkage control routes planned each time as the target number. Or, the average number of used planned coordinated control routes in the planned coordinated control routes at each time is used as the target number.

Fig. 3 illustrates a hardware structural diagram of a cloud service platform 100 for implementing the above-mentioned unmanned aerial vehicle-based intelligent positioning method for a muck truck, according to an embodiment of the present invention, as shown in fig. 3, the cloud service platform 100 may include a processor 110, a machine-readable storage medium 120, a bus 130, and a communication unit 140.

In a specific implementation process, the processors 110 execute the computer executable instructions stored in the machine readable storage medium 120, so that the processors 110 may execute the method for intelligently positioning a muck truck based on a drone according to the above method embodiment, the processors 110, the machine readable storage medium 120, and the communication unit 140 are connected through the bus 130, and the processors 110 may be configured to control the transceiving action of the communication unit 140, so as to perform data transceiving with the monitoring drone 200.

Machine-readable storage medium 120 may store data and/or instructions. In one possible design approach, the machine-readable storage medium 120 may store data and/or instructions used by or to perform the exemplary methods described in this disclosure by the cloud service platform 100. In one possible design approach, the machine-readable storage medium 120 may include mass storage, removable storage, volatile read-write memory, read-only memory (ROM), and the like, or any combination thereof. Exemplary mass storage devices may include magnetic disks, optical disks, solid state disks, and the like. Exemplary removable memory may include flash drives, floppy disks, optical disks, memory cards, compact disks, magnetic tape, and the like. Exemplary volatile read and write memories can include Random Access Memory (RAM). Exemplary RAM may include healthy random access memory (DRAM), double data rate synchronous healthy random access memory (DDR SDRAM), Static Random Access Memory (SRAM), thyristor random access memory (T-RAM), and zero capacitance random access memory (Z-RAM), among others. Exemplary read-only memories may include mask read-only memory (MROM), programmable read-only memory (PROM), erasable programmable read-only memory (perrom), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM), digital versatile disc read-only memory, and the like. In one possible design approach, the machine-readable storage medium 120 may be implemented on the cloud service platform 100. By way of example only, the cloud service platform 100 may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-tiered cloud, and the like, or any combination thereof.

For a specific implementation process of the processor 110, reference may be made to the above-mentioned method embodiments executed by the cloud service platform 100, and implementation principles and technical effects are similar, which are not described herein again.

In addition, the embodiment of the invention also provides a readable storage medium, wherein the readable storage medium is preset with computer executable instructions, and when a processor executes the computer executable instructions, the intelligent positioning method of the muck vehicle based on the unmanned aerial vehicle is realized.

It should be understood that the foregoing description is for purposes of illustration only and is not intended to limit the scope of the present disclosure. Many modifications and variations will be apparent to those of ordinary skill in the art in light of the description of the invention. However, such modifications and variations do not depart from the scope of the present invention.

While the basic concepts have been described above, it will be apparent to those of ordinary skill in the art in view of this disclosure that the above disclosure is intended to be exemplary only and is not intended to limit the invention. Various modifications, improvements and optimization of the invention will occur to those skilled in the art, although not explicitly described herein. Such modifications, improvements and optimized derivatives are proposed in the present invention and, therefore, fall within the spirit and scope of the exemplary embodiments of the present invention.

Also, the present invention has been described using specific terms to describe embodiments of the invention. For example, "one embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic described in connection with at least one embodiment of the invention. Therefore, it is emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, some of the features, structures, or characteristics of one or more embodiments of the present invention may be combined as suitable.

Moreover, those skilled in the art will recognize that aspects of the present invention may be illustrated and described in terms of several patentable species or situations, including any new and useful process, machine, article, or material combination, or any new and useful modification thereof. Accordingly, aspects of the present invention may be embodied entirely in hardware, entirely in software (including firmware, resident software, micro-code, etc.) or in a combination of hardware and software. The above hardware or software may be referred to as a "unit", "module", or "system". Furthermore, aspects of the present disclosure may take the form of a computer program product embodied in one or more computer-readable media, with computer-readable program code embodied therein.

A computer readable signal medium may comprise a propagated data signal with computer program code embodied therein, for example, on a baseband or as part of a carrier wave. Such a propagated signal may take any of a variety of forms, including electro-magnetic, optical, and the like, or any suitable combination. A computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code on a computer readable signal medium may be propagated over any suitable medium, including radio, electrical cable, fiber optic cable, RF, or the like, or any combination thereof.

Computer program code required for the operation of various parts of the present invention may be written in any one or more programming languages, including a subject oriented programming language such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C + +, C #, VB.NET, Python, etc., a conventional programming language such as C, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, a health programming language such as Python, Ruby, and Groovy, or other programming languages, etc. The program code may run entirely on the target drone computer, or as a stand-alone software package, or partly on the target drone computer and partly on a remote computer, or entirely on the remote computer or server. In the latter case, the remote computer may be connected to the target drone computer by any form of network, such as a Local Area Network (LAN) or a Wide Area Network (WAN), or to an external computer (e.g., via the internet), or in a cloud computing environment, or as a service using, for example, software as a service (SaaS).

Additionally, the order in which the elements and sequences of the process are described, the use of letters or other designations herein is not intended to limit the order of the processes and methods of the invention unless otherwise indicated by the claims. While various presently contemplated embodiments of the invention have been discussed in the foregoing disclosure by way of example, it is to be understood that such detail is solely for that purpose and that the appended claims are not limited to the disclosed embodiments, but, on the contrary, are intended to cover all modifications, equivalents, and combinations that are within the spirit and scope of the embodiments of the invention. For example, although the system components described above may be implemented by hardware devices, they may also be implemented by software-only solutions, such as installing the described system on an existing server or mobile device.

Similarly, it should be noted that in the preceding description of embodiments of the invention, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure aiding in the understanding of one or more of the embodiments. Similarly, it should be noted that in the preceding description of embodiments of the invention, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure aiding in the understanding of one or more of the embodiments.

Claims

1. An intelligent muck vehicle positioning method based on an unmanned aerial vehicle is applied to a cloud service platform, wherein the cloud service platform is in communication connection with a plurality of monitoring unmanned aerial vehicles, and the method comprises the following steps:

when the monitoring data stream is detected to indicate that the target muck truck has violation behaviors, other unmanned aerial vehicles in the target control area are controlled in a linkage mode to acquire driving path information of the target muck truck in real time; acquiring flight state information of all candidate unmanned aerial vehicles in the target control area;

screening all candidate unmanned aerial vehicles according to the flight state information to obtain target unmanned aerial vehicles, and generating corresponding linkage control strategy information for each target unmanned aerial vehicle according to the current driving direction information of the target muck vehicle and the flight state information of each target unmanned aerial vehicle;

controlling each corresponding target unmanned aerial vehicle to acquire driving path information of the target muck truck in real time based on each corresponding linkage control strategy information;

2. The intelligent unmanned aerial vehicle-based muck vehicle positioning method according to claim 1, wherein the step of predicting the driving positioning prediction information of the target muck vehicle within a preset time period after the target muck vehicle is predicted based on the driving path information of the target muck vehicle collected by each unmanned aerial vehicle comprises the following steps:

extracting a driving path characteristic sequence of the driving path information of the target muck truck acquired by each unmanned aerial vehicle, and fusing the driving path characteristic sequence of the driving path information with a corresponding flight state characteristic sequence of the unmanned aerial vehicle respectively to obtain a fused path characteristic cluster formed by each fused path characteristic sequence;

inputting the fusion path feature cluster into a driving positioning prediction network obtained by pre-training, and predicting driving positioning prediction information of the target muck vehicle in a later preset time period;

wherein the method further comprises:

acquiring a sample fusion path feature cluster and sample driving positioning prediction information corresponding to the sample fusion path feature cluster, and performing path feature description splitting on the sample fusion path feature cluster to obtain a path feature description set corresponding to driving path track information; the path feature description set corresponding to the driving path track information comprises a plurality of path feature descriptions;

carrying out path mutation node analysis on a path characteristic description set corresponding to the driving path track information to obtain different path characteristic descriptions, and training a driving location prediction network through an input data sequence formed by the different path characteristic descriptions and the sample driving location prediction information; the driving positioning prediction network is a pre-configured AI neural network model, and the trained driving positioning prediction network is used for driving positioning prediction; and the driving positioning prediction network is used for performing driving positioning prediction on the fusion path feature cluster.

3. The intelligent positioning method for the muck vehicle based on the unmanned aerial vehicle as claimed in claim 2, wherein the step of splitting the path feature description of the sample fusion path feature cluster to obtain a path feature description set corresponding to the driving path trajectory information comprises the steps of:

for each group of sample fusion path feature clusters in the traffic path track information, performing path feature description splitting on the traffic path track information of the sample fusion path feature clusters according to corresponding path track catastrophe points to obtain a path feature description set corresponding to the traffic path track information of the sample fusion path feature clusters;

carrying out path mutation node analysis on a path feature description set corresponding to the driving path track information to obtain different path feature descriptions, and training a driving location prediction network through an input data sequence formed by the different path feature descriptions and the sample driving location prediction information, wherein the method comprises the following steps:

for a plurality of path feature descriptions included in a path feature description set corresponding to the path trace information of each group of sample fusion path feature clusters, determining the path feature description of which the path azimuth change times exceed a target time threshold value as a path feature description set corresponding to the path trace information of the undetermined vehicle;

performing key feature matching processing on the path feature description in the path feature description set corresponding to the to-be-determined driving path track information, so that the path feature description meeting the key feature matching condition is used as the expressive path feature description for representing the path track floating change of the driving path track information;

determining a plurality of non-expressive path profiles, wherein the non-expressive path profiles are path profiles contained in path profiles of which the path azimuth change times are not more than the target time threshold;

screening part of the non-expressive path profiles as reference path profiles according to related profiles among the non-expressive path profiles in a plurality of the non-expressive path profiles;

and training a driving location prediction network based on the expressive path characteristic description, the reference path characteristic description and the sample driving location prediction information of each group of sample fusion path characteristic clusters in the driving path track information.

4. The intelligent positioning method for the muck vehicle based on the unmanned aerial vehicle as claimed in claim 3, wherein the step of splitting the path description of the traffic path trajectory information of the sample fused path feature cluster according to the corresponding path trajectory mutation point to obtain a path feature description set corresponding to the traffic path trajectory information of the sample fused path feature cluster comprises:

extracting path mutation characteristics of dynamic path track mutation points and static path track mutation points from the path mutation area information of the sample fusion path characteristic cluster, and generating a plurality of driving path state transition characteristics based on the extracted path mutation characteristics;

according to the state transition labels of each group of the driving path state transition characteristics, determining a target decomposition rule of the state transition labels based on a track tracking decomposition rule of corresponding path track break points;

decomposing the traffic path track information of the sample fusion path feature cluster according to a target decomposition rule determined by a state transition label corresponding to each group of traffic path state transition features to obtain a path feature description set corresponding to the traffic path track information of the sample fusion path feature cluster;

before determining a target decomposition rule of the state transition label according to the state transition label of each group of the driving path state transition characteristics and based on a track tracking decomposition rule of a corresponding path track break point, the method further includes a preprocessing step performed on each group of the driving path state transition characteristics in the plurality of driving path state transition characteristics and the corresponding state transition label, where the preprocessing step includes:

determining the state transition times of the state transition labels of each group of driving path state transition characteristics and the statistic value of the same state transition label of the same driving path state transition characteristics;

cleaning the traffic route state transition characteristics of which the state transition times of the state transition labels exceed a preset transition time threshold and the statistical value of the same state transition labels exceed a preset statistical threshold corresponding to the preset transition time threshold so as to obtain the remaining traffic route track information after cleaning;

filtering noise state transition labels corresponding to the rest of the cleaned path track information, and updating state transition labels with abnormal state transition in the rest of the cleaned path track information;

and obtaining the driving path state transfer characteristics for path characteristic description splitting based on the rest driving path track information after cleaning and the updated state transfer label.

5. The intelligent positioning method for the muck vehicle based on the unmanned aerial vehicle as claimed in claim 3, wherein the step of determining, for a plurality of path profiles included in the path profile set corresponding to the travel path trajectory information of each group of the sample fused path profile clusters, a path profile whose number of changes in the direction of the path exceeds a target number threshold as the path profile set corresponding to the path trajectory information of the undetermined vehicle comprises:

screening the path feature description for multiple times in path feature description sets corresponding to multiple driving path track information of different sample fusion path feature clusters, and gathering and splicing the multiple path feature descriptions screened each time to obtain multiple different path feature description sets; the method comprises the steps that path feature descriptions in a path feature description set corresponding to driving path track information of different sample fusion path feature clusters are covered in a plurality of path feature descriptions screened each time;

and screening the path feature description set with the path azimuth change times exceeding the target time threshold value in a path feature description group formed by the different path feature description sets to serve as a path feature description set corresponding to the path track information of the undetermined vehicle.

6. The intelligent positioning method for the muck vehicle based on the unmanned aerial vehicle as claimed in claim 3, wherein the key feature matching processing is performed on the path feature descriptions in the path feature description set corresponding to the path trajectory information of the vehicle to be determined, and comprises the following steps:

extracting a plurality of path feature descriptions from a path feature description set corresponding to the path track information of the undetermined vehicle;

executing at least one of the following processing modes on the extracted multiple path feature descriptions:

performing key feature matching on local feature descriptions in the path feature descriptions to obtain a key feature matching result aiming at the local feature descriptions;

performing key feature matching on an extended feature vector set of feature vectors corresponding to each group of path feature descriptions in the plurality of path feature descriptions to obtain a key feature matching result for the extended feature vector set;

respectively adding time-space domain feature marks in the path feature descriptions to perform key feature matching to obtain key feature matching results aiming at the time-space domain features;

performing key feature matching on the path feature descriptions according to a shared feature description part among the path feature descriptions to obtain a key feature matching result aiming at the shared feature description;

and if the key confidence corresponding to at least one key feature matching result is greater than the preset confidence, determining the path feature description corresponding to the at least one key feature matching result as the path feature description meeting the key feature matching condition.

7. The unmanned-aerial-vehicle-based intelligent muck vehicle positioning method of claim 3, wherein the driving location prediction network comprises a plurality of driving location analysis subnetworks; the training of the driving location prediction network based on the expressive path feature description, the reference path feature description and the sample driving location prediction information of each group of sample fusion path feature clusters in the driving path trajectory information includes:

combining the expressive path characteristic description, the reference path characteristic description and the sample driving positioning prediction information of each group of sample fusion path characteristic clusters in the driving path track information into a training data sequence;

training the plurality of driving location analysis sub-networks based on the composed plurality of training data sequences;

and cascading the trained plurality of driving positioning analysis sub-networks through the driving positioning analysis dimensionality to obtain the driving positioning prediction network.

8. The intelligent positioning method for the muck vehicle based on the unmanned aerial vehicle as claimed in claim 7, wherein before the step of performing path feature description splitting on the travel path trajectory information of the sample fusion path feature cluster according to the corresponding path trajectory mutation point for each group of sample fusion path feature clusters in the travel path trajectory information to obtain the path feature description set corresponding to the travel path trajectory information of the sample fusion path feature cluster, the method further comprises:

acquiring path mutation area information of a plurality of previous fusion path characteristic information;

according to prediction feedback information of sample driving positioning prediction information, carrying out path mutation quality analysis on the path mutation area information of the plurality of previous fusion path characteristic information to obtain evaluation information of each group of previous fusion path characteristic information, wherein the evaluation information is used for screening the sample fusion path characteristic cluster from the previous fusion path characteristic information;

screening partial previous fusion path characteristic information as a sample fusion path characteristic cluster according to the evaluation information of the plurality of previous fusion path characteristic information;

wherein, the screening of partial previous fusion path feature information as a sample fusion path feature cluster according to the evaluation information of the plurality of previous fusion path feature information includes at least one of:

screening the fusion path characteristic information before the part of which the evaluation metric value corresponding to the evaluation information is higher than the preset evaluation metric value in the plurality of previous fusion path characteristic information to be used as a sample fusion path characteristic cluster;

according to the evaluation metric values corresponding to the evaluation information of the plurality of previous fusion path feature information, descending the plurality of previous fusion path feature information, and screening the fusion path feature information which is ranked in the front and is before the part with the number of the set number to be used as a sample fusion path feature cluster;

wherein, in the plurality of non-representational path feature descriptions, screening part of the non-representational path feature descriptions as reference path feature descriptions according to relevant feature descriptions among the non-representational path feature descriptions comprises:

determining the validity times of the plurality of non-expressive path characteristic descriptions, and cleaning the non-expressive path characteristic descriptions with the validity times lower than the preset times;

collecting and splicing the rest non-expressive path characteristic descriptions after cleaning to obtain a non-expressive path characteristic description set;

determining a relevant feature description between two random groups of non-representational path feature descriptions in the set of non-representational path feature descriptions;

determining a local correlation feature description of each non-representational path feature description in the non-representational path feature description set and the non-representational path feature description set according to the correlation feature description between the random two sets of non-representational path feature descriptions;

and according to the local relevant feature descriptions, descending the non-expressive path feature descriptions in the non-expressive path feature description set according to the content relevance corresponding to the local relevant feature descriptions, and screening the part of the non-expressive path feature descriptions which are ranked in the front as reference path feature descriptions.

9. An intelligent muck vehicle positioning system based on an unmanned aerial vehicle is characterized by comprising a cloud service platform and a plurality of monitoring unmanned aerial vehicles in communication connection with the cloud service platform;

the cloud service platform is used for: