Intelligent illumination inspection method based on multi-source data perception and inspection robot system

Abstract

The invention discloses an intelligent illumination inspection method based on multi-source data perception and an inspection robot system, and the method comprises the steps: dividing a road which needs to be subjected to illumination detection according to the positions of street lamps, taking a part between every two adjacent street lamps as a road section, and sequentially carrying out the illumination detection of all road sections through an inspection robot, when one road section is detected, all road sections divided by the inspection robot are subjected to illumination detection in sequence, a to-be-detected operation area is divided by positioning the positions of lamp poles of street lamps, then an illumination detection grid array is divided, detection points distributed according to a matrix are obtained, and then all the detection points are detected by the inspection robot in sequence. Therefore, comprehensive inspection of the road section is completed. The inspection robot system realizes automation of illumination inspection through an autonomous navigation vehicle body, a depth camera, a multi-line laser radar, a spectrometer and the like, so that manual detection is replaced, the detection efficiency is improved, the operation and maintenance cost can be saved, and the personal safety of a detector is better ensured.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of automatic inspection of road lighting, and particularly relates to an intelligent illumination inspection method based on multi-source data perception and an inspection robot system. BACKGROUND

[0002] Functional lighting plays a prominent role in facilitating citizens' night travel and improving people's well-being. For example, in Guangdong Province, Guangdong actively promotes the green and low-carbon development of urban lighting. According to the statistics of the Housing and Urban-Rural Development Department of Guangdong Province, there are 4.2334 million functional lighting in the province in 2024, and the number of LED lamps accounts for 84.26%; there are more than 9.3916 million landscape lighting, and the number of LED lamps accounts for 59.46%. The energy-saving reconstruction of urban lighting has achieved initial success.

[0003] The basis of energy-saving reconstruction of urban lighting is to detect and obtain the relevant illumination parameters of functional lighting on the road. In the existing technical solutions, the detection of illumination parameters of urban functional lighting can only be carried out by manual method. The detector divides the detection area on the road according to the lighting detection standard (such as "Lighting Measurement Method" GB / T5700-2023 and "Road Lighting Engineering Technical Specification" DBJ / T 15-242-2022), and then places the instrument at the specified measurement point for detection by the detector carrying the illumination detector. After detection, the data is recorded manually on the form.

[0004] The above-mentioned manual detection method has the following key problems:

[0005] (1) The operation efficiency of manual detection is low, a large amount of manpower and material resources are consumed, and the road lamp faults cannot be effectively known from the technical level;

[0006] (2) The outdoor lighting detection environment is complex, the safety risk of the detector on the road is high, the personal safety is affected by the road conditions, the detection safety hidden danger is prominent, and there is a large traffic safety hidden danger;

[0007] (3) The types and working conditions of lighting equipment are various, and the operation and maintenance cost is huge;

[0008] (4) It is not conducive to the unified operation and maintenance management of urban functional lighting, and the fault points cannot be found in time. SUMMARY

[0009] The first object of the present application is to provide an intelligent illumination inspection method based on multi-source data perception.

[0010] The second object of the present application is to provide an inspection robot system applied to the intelligent illumination inspection method.

[0011] The first object of the present application is achieved by the following technical solutions:

[0012] An intelligent illumination inspection method based on multi-source data sensing, characterized in that: the road to be subjected to illumination detection is divided according to the positions of the street lamps, each adjacent two street lamps serving as a road section, and all the road sections are subjected to illumination detection in turn by an inspection robot;

[0013] The steps of illumination detection on a road section are as follows:

[0014] S1, initialization of the posture and position of the inspection robot;

[0015] S2, the inspection robot photographs and recognizes the street lamps, locates the positions of the lamp poles of the two street lamps, obtains a quadrilateral to-be-detected work area extending along the road according to the positions of the lamp poles, the positions of the two lamp poles corresponding to the two ends of the to-be-detected work area, divides an illumination detection grid array in the to-be-detected work area, and takes the geometric center of each grid of the illumination detection grid array as a detection point;

[0016] Meanwhile, the inspection robot scans the surrounding environment and constructs a surrounding environment sensing model;

[0017] S3, the inspection robot autonomously navigates to the detection points according to the constructed surrounding environment sensing model and coordinates, detects all the detection points in turn by a spectrometer on the inspection robot, and respectively obtains optical characteristic parameters.

[0018] A further technical solution of the present application is that: in step S3, after detecting each detection point, it is judged whether the obtained optical characteristic parameters are complete, if not, the position is confirmed again and detected again, if yes, the next detection point is navigated to for detection; when all the detection points of a road section are detected, it is judged whether the detection of the optical characteristic parameters of all the detection points is completed, if not, the corresponding detection point is re-detected by re-navigation, if yes, the initial position is returned, and the optical characteristic parameters are returned to a remote server at the same time.

[0019] A further technical solution of the present application is that: after step S2 is completed, the detection task of the to-be-detected road section is reviewed by a detector, and then step S3 is continued after determination.

[0020] The second object of the present application is achieved by the following technical solution:

[0021] The application discloses an intelligent illumination inspection method and a robot system thereof.

[0022] The application further provides a technical scheme that the remote server comprises a multi-source perception database, a job operation database and a historical parameter database.

[0023] The application further provides a technical scheme that the hardware part of the robot further comprises a display used for displaying the results processed by the visual recognition module, the mapping navigation module and the illumination detection module.

[0024] The application further provides a technical scheme that the robot further comprises an infrared transmitting-receiving automatic recharging device, and the automatic recharging of the robot under low power is realized through the infrared transmitting-receiving automatic recharging device.

[0025] Compared with the prior art, the application has the following beneficial effects:

[0026] The intelligent illumination inspection method of the present application carries out illumination detection on all road sections divided according to the positions of street lamps in sequence by an inspection robot, which divides the to-be-detected work area by positioning the positions of the lamp posts of the street lamps, and further divides the illumination detection grid array to obtain the detection points distributed in a matrix, and then the inspection robot sequentially detects all the detection points to complete the comprehensive inspection of the road section. The inspection robot system realizes the automation of illumination inspection by means of the self-navigating vehicle body, the depth camera, the multi-line laser radar, the spectrometer and the like, thereby replacing manual detection, improving the detection efficiency, saving the operation and maintenance cost, and better ensuring the personal safety of the inspectors. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 Fig. 1 is a structural schematic diagram of the inspection robot of an embodiment of the present application;

[0028] Figure 2 Fig. 2 is a structural schematic diagram of the inspection robot of another embodiment of the present application;

[0029] Figure 3 Fig. 3 is a flow chart of the illumination detection of a road section by the intelligent illumination inspection method of an embodiment of the present application;

[0030] Figure 4 Fig. 4 is a schematic diagram of the illumination detection of a road section in the field by the intelligent illumination inspection method of an embodiment of the present application;

[0031] Figure 5 Fig. 5 is a data flow processing diagram of the multi-source data sensing in the inspection robot system of an embodiment of the present application;

[0032] Figure 6 Fig. 6 is an architectural diagram of the inspection robot system of an embodiment of the present application;

[0033] Figure 7 Fig. 7 is a comparison diagram of the parameter accuracy of the illumination detection by the inspection robot of the present application and the illumination detection by the traditional manual method on the same road;

[0034] Figure 8 Fig. 8 is a comparison diagram of the detection efficiency of the illumination detection by the inspection robot of the present application and the illumination detection by the traditional manual method on the same road.

[0035] Meaning of reference signs in the drawings:

[0036] 1-depth camera; 2-multi-line laser radar; 3-core computer; 4-spectrometer; 5-self-navigating vehicle body; 6-display; 7-infrared transceiver automatic recharging device; 8-power brake button. DETAILED DESCRIPTION

[0037] The present application is further described below in combination with embodiments.

[0038] Embodiment:

[0039] The intelligent illumination inspection method based on multi-source data sensing of the embodiment is to divide the road needing illumination detection according to the positions of the street lamps, take each two adjacent street lamps as a road section, and sequentially perform illumination detection on all road sections by the inspection robot, so that the inspection of the whole road can be finally completed.

[0040] The steps of performing illumination detection on a road section refer to Figure 3 , and are as follows:

[0041] S1, initialization of the inspection robot, specifically, initialization of the robot posture and the position where the robot is located, obtaining the current relevant posture and position parameters, and self-checking the online running of each function to ensure the stability of the illumination parameter detection operation.

[0042] S2, after the initialization is performed, entering the area calculation step, shooting the road environment by the inspection robot, recognizing the street lamps, positioning the positions of the lamp poles of the two street lamps, calculating a quadrilateral to-be-detected operation area extending along the road according to the positions of the lamp poles, and corresponding the positions of the two lamp poles to the two ends of the to-be-detected operation area; dividing an illumination detection grid array in the to-be-detected operation area by coordinate conversion, and taking the geometric center of each grid of the illumination detection grid array as a detection point K;

[0043] As Figure 4 shown is a real scene diagram of performing illumination detection on a road section, the to-be-detected operation area (referring to the to-be-detected area in Figure 4 ) is divided into an illumination detection grid array of three rows and ten columns, having thirty grids, that is, thirty detection points K;

[0044] At the same time, the inspection robot scans the surrounding environment to construct a surrounding environment sensing model for autonomous navigation.

[0045] After step S2 is completed, the detection task of the to-be-detected road section is reviewed by a detection operator, and after the execution operation instruction is determined to be issued, step S3 is continued to be executed.

[0046] S3, after receiving the execution operation instruction, entering the detection operation, the inspection robot autonomously navigates to the detection points according to the constructed surrounding environment sensing model; first, autonomously navigating to a detection point, after reaching the detection point, triggering the spectrometer on the inspection robot to perform detection, obtaining the light characteristic parameters of the first detection point, and judging whether the obtained light characteristic parameters are complete (it can be a logical judgment whether the light characteristic parameters are empty), if not, reconfirming the position and re-detecting, if yes, navigating to the next detection point for detection; repeating the above steps to finally complete the detection of all detection points of the to-be-detected road section.

[0047] As Figure 4 shown in the illumination detection grid array, the upper left corner is the first checkpoint, and the detection sequence of the inspection robot on the detection point is shown by the arrow in the figure, forming an S-shaped route.

[0048] When all the detection points of a section are detected, it is determined whether the light characteristic parameters of all the detection points are completed (it can be a logical judgment on whether the light characteristic parameters of all the detection points are empty in sequence), if not, then re-navigate to the corresponding detection point with empty light characteristic parameters for re-detection, if yes, then return to the initial position and return the light characteristic parameters to the remote server. The light characteristic parameters returned to the remote server will be stored for data original record and converted to generate an illumination parameter detection report.

[0049] In this embodiment, the above-mentioned intelligent illumination inspection method is implemented by the following inspection robot system:

[0050] The inspection robot system includes an inspection robot, a remote server in communication with the inspection robot, and a user terminal in communication with the remote server.

[0051] As Figure 1 and Figure 2 shown, the hardware part of the inspection robot includes an autonomous navigation vehicle body 5, a depth camera 1, a multi-line laser radar 2, a spectrometer 4, a core computer 3, a display 6, an infrared transmitting and receiving automatic charging device 7, and a communication device for communication with the remote server. The autonomous navigation vehicle body 5, the depth camera 1, the multi-line laser radar 2, the spectrometer 4, the display 6, the infrared transmitting and receiving automatic charging device 7, and the communication device are all electrically connected to the core computer 3.

[0052] Specifically, the depth camera 1 adopts an infrared enhanced binocular depth camera, which automatically switches to an infrared enhancement mode in poor light conditions or dim environments, and is suitable for various light conditions. When in use, the depth camera 1 captures the road environment, and the video image data captured by the depth camera 1 can be used for the recognition and positioning of the street lamp pole. The multi-line laser radar 2 adopts a 32-line omnidirectional multi-line laser radar, which can realize perception and reconstruction of the environment around the inspection robot by combining TOF inversion calculation, form an environmental model around the robot, and thus realize the functions of positioning, navigation, intelligent obstacle avoidance, etc. in the detection of the robot. The spectrometer 4 adopts a first-order miniature spectrometer, which receives the light parameters in the environment through an optical probe, detects various light characteristic parameters such as full-spectrum distribution, illumination, brightness, and color temperature of the light in the environment, and transmits the parameters back to the core calculator through serial communication. The core calculator is equipped with a Jetson AI development and algorithm processing core kit. After receiving multi-source information data including images, laser radar signals, and illumination detection parameters, the core calculator performs synchronous parallel processing and outputs corresponding data processing results (such as image recognition results, environmental reconstruction models, and illumination integral parameters). The display is used to display the corresponding results processed by the core calculator, including image recognition results, environmental perception modeling, and illumination integral parameters. The autonomous navigation vehicle body 5 adopts a four-wheel independent suspension power system, which adopts four independent suspensions and independent brushless motors, has full-terrain adaptability such as climbing and passing through pits, and can move and pass through various urban lighting roads. The infrared transmitting and receiving automatic charging device 7 adopts existing automatic charging technology, mainly including an infrared signal transmitting and receiving module and a contact electrode, which positions the charging base through the infrared transmitting and receiving mode, triggers automatic charging in the case of low power, and realizes the rapid charging of the inspection robot through the contact electrode.

[0053] The embodiment further provides a power brake button 8 on the inspection robot, which is used to brake the power of the inspection robot in complex situations, disconnect the power output of the inspection robot, and stop the inspection robot.

[0054] The software part of the inspection robot includes a visual recognition module, a mapping navigation module, and an illumination detection module. The visual recognition module is used to obtain video image data of the depth camera 1 and process it, locate the positions of the lamp posts of the two street lamps, obtain the to-be-detected work area, and divide the illumination detection grid array through coordinate conversion. It mainly uses existing image feature recognition algorithms and three-dimensional coordinate conversion algorithms. The mapping navigation module is used to obtain radar reconstruction model data scanned by the multi-line laser radar 2 and process it, and finally construct a surrounding environment perception model for autonomous navigation. It mainly uses existing radar reconstruction algorithms. The illumination detection module is used to obtain spectral data detected by the spectrometer 4 and process it to obtain corresponding light characteristic parameters. It mainly uses existing illumination detection integration algorithms. The image feature recognition algorithm, the radar reconstruction algorithm, the three-dimensional coordinate conversion algorithm, and the illumination detection integration algorithm process the video image data collected by the binocular depth camera 1, the inversion data of the multi-line laser radar 2, the to-be-detected region and the robot pose relationship calculation data, and the real-time illumination detection data, respectively, to realize the complete inspection robot illumination automatic detection work function.

[0055] As Figure 5 The data flow processing diagram of multi-source data perception in the inspection robot system is shown. The hardware and software parts on the inspection robot are the edge computing end. The omnidirectional multi-line laser radar, the first micro spectrometer, the binocular depth camera, and the infrared transceiver module carried on the inspection robot are the perception hardware of multi-source information data. Radar reconstruction model data, lighting spectral parameters (illumination), video image data, and infrared signal transmission data are obtained through the perception hardware. The remote server includes a multi-source perception database, a work operation database, and a historical parameter database. The original data output by the perception hardware is immediately sent to the multi-source perception database of the remote server for storage. During the detection process of the road section, different multi-source data are processed in the edge computing end. The data fields in the work operation database are synchronized to the historical parameter database for storage and record at the end of each detection period, which is convenient for management and retrieval of historical detection data. The databases of the remote server can be accessed and managed by the user end through Ethernet. The user end can be a PC client and a portable client.

[0056] To more specifically present the inspection robot system, the inspection robot system will be further described in five levels as follows. Figure 6As shown, it is specifically divided into equipment layer constituting system base, link layer of data transmission path, data layer of data storage management, middle layer of core processing algorithm calculation and function implementation, and application layer facing users, each level is correlated and expanded from top to bottom, and deepens the implementation of various functions of intelligent illumination inspection.

[0057] The equipment layer is composed of three hardware blocks of information sensing, core calculator and communication device, wherein the information sensing is composed of multi-source information sensing hardware, i.e. including binocular depth camera, omnidirectional multi-line laser radar, first-level miniature spectrometer and infrared transceiver module, and the above multi-source sensing information is transferred and integrated. The core calculator is a calculation hardware carried on the robot body, containing a main control module and a calculation processing module, used for running multi-source information core processing algorithm, and additionally configured with an SSD for multi-source information processing algorithm configuration storage and multi-source data temporary cache. The communication device contains a wireless network adaptation terminal and a firewall, respectively used for data receiving and sending and link security protection.

[0058] The link layer is composed of wired link, physical interface and transmission protocol, the inspection robot, remote server and user end adopt the Ethernet port form of TCP / IP for data stream transmission; and various sensors carried on the inspection robot transmit data through serial communication such as RS-485 and USB.

[0059] The data layer is composed of different database modules, respectively multi-source sensing database, operation running database and historical parameter database. Different databases are carried in the remote server, designed according to data characteristics, and transmitted, synchronized and managed through Ethernet.

[0060] The middle layer is composed of core processing algorithm and management module, the core processing algorithm module is composed of four algorithm modules of image feature recognition algorithm, radar reconstruction algorithm, three-dimensional coordinate conversion algorithm and illumination detection integral algorithm, respectively used for processing video image information collected by the depth camera, multi-line laser radar inversion data, to-be-detected area and robot posture relationship calculation data, and real-time illumination detection data, so as to realize the complete robot illumination automatic detection operation function. In addition, the management module includes four program modules of data management, system log, communication service and user management for daily system platform management.

[0061] The application layer is composed of two blocks of online state monitoring and data conversion and management, mainly related to functional application modules of the system platform, the online state monitoring mainly includes real-time image display, reconstructed environment model and automatic navigation route generation, detection point detection illumination parameter monitoring, etc., the data conversion and management mainly includes detection parameter conversion to generate report, original data record and historical parameter query, etc., in the way of user interface for detection personnel to operate and monitor.

[0062] In practical applications, the intelligent illumination inspection method and inspection robot system of this invention can greatly improve inspection efficiency. For example... Figure 7 The figure shows a comparison of the accuracy of illuminance detection parameters performed by an inspection robot and by traditional manual methods on the same road. As can be seen from the figure, the accuracy of the inspection robot and traditional manual methods is very similar; the distribution, trend, and values ​​of the illuminance parameters are basically consistent. This demonstrates that using the inspection robot of this invention for illuminance inspection does not affect the detection accuracy. Figure 8 The figure shows a comparison of the detection efficiency of illuminance detection by inspection robots and traditional manual methods on the same road. As can be seen from the figure, the detection efficiency of inspection robots remains relatively stable, fluctuating around 180 points per hour depending on the terrain. However, with manual detection, the efficiency drops sharply as the continuous working time increases due to fatigue of the inspectors.

[0063] From the above Figure 7 and Figure 8 The analysis shows that the inspection robot can greatly improve the inspection efficiency without affecting the accuracy of the illumination parameter detection. The efficiency improvement exceeds 50% when the number of inspection points is used as the evaluation.

[0064] In addition to improved accuracy and efficiency, using inspection robots can largely mitigate the risks associated with manual inspections on highways. Inspectors can remotely monitor from a safe area and intervene manually when necessary, thus greatly reducing the risk of unexpected situations during urban lighting inspections.

[0065] By conducting regular inspections with inspection robots in conjunction with urban lighting monitoring systems, managers can gain a timely understanding of the status of urban lighting equipment in different areas and at different times. This allows for the timely detection of malfunctions and prompt maintenance of the lighting equipment. On the one hand, it reduces the duration of malfunctions, and on the other hand, it effectively improves the level of urban road lighting and enhances the city's image.

[0066] The above embodiments of the present invention are not intended to limit the scope of protection of the present invention. The implementation of the present invention is not limited thereto. All other modifications, substitutions or alterations made to the above structure of the present invention based on the above content of the present invention, in accordance with ordinary technical knowledge and common practice in the field, without departing from the basic technical idea of ​​the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A smart illumination inspection method based on multi-source data sensing, characterized in that: The roads that need to be tested for illuminance are divided according to the location of the streetlights. Each section is defined as the area between two adjacent streetlights. The inspection robot then conducts illuminance testing on all sections in sequence. The steps for conducting illuminance testing on a road section are as follows: S1, Initialize the posture and position of the inspection robot; S2, the inspection robot takes pictures and identifies streetlights, locates the positions of the two streetlight poles, and obtains a quadrilateral inspection area extending along the road based on the position of the poles. The positions of the two poles correspond to the two ends of the inspection area. An illuminance detection grid array is divided in the inspection area, and the geometric center of each grid of the illuminance detection grid array is used as the detection point. At the same time, the inspection robot scans the surrounding environment and builds a perception model of the surrounding environment; S3, the inspection robot autonomously navigates to the detection point based on the constructed surrounding environment perception model and coordinates, and uses the spectrometer on the inspection robot to detect all the detection points in sequence to obtain the optical characteristic parameters of each point.

2. The intelligent illumination inspection method based on multi-source data perception according to claim 1, characterized in that: In step S3, after each detection point is detected, it is determined whether the obtained optical characteristic parameters are complete. If not, the position is reconfirmed and re-detected. If it is complete, navigation is performed to the next detection point for detection. Once all detection points for a road segment have been detected, determine whether the optical characteristic parameters of all detection points have been detected. If not, re-navigate to the corresponding detection point for re-detection. If yes, return to the initial position and simultaneously send the optical characteristic parameters back to the remote server.

3. The intelligent illumination inspection method based on multi-source data perception according to claim 1, characterized in that: After completing step S2, the inspector will first review the inspection task of the road section to be inspected, and then proceed to step S3 after confirmation.

4. An inspection robot system used in the intelligent illumination inspection method according to any one of claims 1 to 3, characterized in that: The inspection robot system includes an inspection robot, a remote server communicating with the inspection robot, and a user terminal communicating with the remote server. The hardware of the inspection robot includes an autonomous navigation vehicle, a depth camera, a multi-line LiDAR, a spectrometer, a core calculator, and a communication device for communicating with the remote server. The autonomous navigation vehicle, depth camera, multi-line LiDAR, spectrometer, and communication device are all electrically connected to the core calculator. The software of the inspection robot includes a vision recognition module, a mapping and navigation module, and an illuminance detection module. The vision recognition module acquires and processes video image data from the depth camera, locates the positions of the lampposts of two streetlights, obtains the area to be inspected, and divides it into an illuminance detection grid array through coordinate transformation. The mapping and navigation module acquires and processes the scanning data from the multi-line LiDAR to construct a surrounding environment perception model for autonomous navigation. The illuminance detection module acquires and processes the spectral data detected by the spectrometer to obtain the corresponding optical characteristic parameters.

5. The inspection robot system according to claim 4, characterized in that: The remote server includes a multi-source perception database, an operation database, and a historical parameter database. The multi-source perception database is used to store the raw data output by the depth camera, multi-line lidar, and spectrometer. The operation database is used to store the data processed by the visual recognition module, mapping and navigation module, and illuminance detection module in the road segment being detected. After each road segment is detected, the data in the operation database is transferred to the historical parameter database for storage.

6. The inspection robot system according to claim 4, characterized in that: The hardware of the inspection robot also includes a display for showing the results processed by the visual recognition module, the mapping and navigation module, and the illuminance detection module.

7. The inspection robot system according to claim 4, characterized in that: The inspection robot also includes an infrared transceiver automatic recharging device, which enables the inspection robot to automatically recharge when the battery is low.

Images