Intelligent full-scene inspection method for photovoltaic power station and multi-agent robot

Abstract

The invention discloses a photovoltaic power station intelligent full-scene inspection method and a multi-agent robot, and the method comprises the steps: monitoring the state of photovoltaic region equipment through a spectral imaging technology; monitoring whether the photovoltaic area equipment has a fault or not through non-intrusive fault voiceprint recognition; photovoltaic area equipment is monitored through a visual parameter collecting and processing technology; and identifying the instruments and meters in the photovoltaic area through automatic copying. Intelligent diagnosis and fault feature modeling of a photovoltaic power generation unit and in-station power transmission and transformation equipment are researched by means of computer vision multi-scale analysis, non-intrusive fault voiceprint recognition, machine learning algorithm optimization and the like; accurate positioning information of equipment faults is obtained through a multi-position image calibration and satellite navigation technology; according to the technical scheme, automatic inspection and multi-machine collaborative operation research in a complex environment are carried out, and key technical breakthrough of intelligent inspection of the photovoltaic power station is finally obtained through field test verification, so that technical support and practical cases are provided for unmanned / few-person operation and maintenance of the photovoltaic power station.

Description

Technical Field

[0001] The present invention relates to the technical field of inspection robots, and in particular to an intelligent full-scene inspection method for photovoltaic power stations and a multi-agent robot. Background Art

[0002] With the continuous expansion of installed capacity and the increase in the number of equipment, how to effectively conduct inspection and maintenance of photovoltaic equipment to ensure the safe operation of power stations has become an urgent problem to be solved. Inspection is an important means to timely discover defects and hidden dangers. Through regular inspections, the operating status and performance of key equipment can be grasped in time, so that measures can be taken when necessary to avoid the expansion of faults, reduce the rate of non-stop accidents, and ensure the stable operation and power supply of power stations.

[0003] At present, the photovoltaic inspection in the industry is mainly based on manual inspection, drone inspection, and slide-type robot inspection, but the above methods have deficiencies in reliability or performance and cannot independently support the photovoltaic intelligent inspection technology system. Drone inspection cannot be applied to airspace restricted areas, and is greatly affected by weather. It cannot cover the inspection of fine parts such as the junction box and MC4 plug on the back of the component, the incoming and outgoing cables of the converging inverter equipment, and the temperature of the meter of the step-up transformer equipment; although the slide-type robot can realize the inspection of indoor power distribution and automation equipment, the actual application effect is not ideal due to its limitations such as large amount of later maintenance, high installation cost of multiple cabins, and unadjustable travel trajectory; manual inspection obviously has problems such as low efficiency and inadequate hidden danger detection. Therefore, in order to further promote the unmanned / less-manned management of photovoltaic sites, it is urgent to develop a set of efficient and intelligent inspection systems that can cover the entire photovoltaic scene to achieve cost reduction and efficiency improvement. Summary of the invention

[0004] In view of the problems existing in the above-mentioned existing photovoltaic power station intelligent full-scene inspection method, the present invention is proposed.

[0005] Therefore, the object of the present invention is to provide an intelligent full-scene inspection method for photovoltaic power stations.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a photovoltaic power station intelligent full-scene inspection method, comprising:

[0007] Monitor the status of photovoltaic area equipment through spectral imaging technology;

[0008] Monitor whether there are faults in photovoltaic area equipment through non-intrusive fault soundprint recognition;

[0009] Monitoring photovoltaic area equipment through visual parameter acquisition and processing technology; and,

[0010] Identify photovoltaic area instrumentation through automated transcription.

[0011] As a preferred solution of the photovoltaic power station intelligent full-scene inspection method of the present invention, wherein: the monitoring of the status of photovoltaic area equipment by spectral imaging technology refers to using spectral imaging technology to provide a basis for judging the health status of photovoltaic area equipment; including;

[0012] Collecting image data;

[0013] Establish a spectral feature database of photovoltaic area equipment through an abnormal feature recognition algorithm;

[0014] Based on machine learning and deep learning algorithms, the model is trained using labeled feature images to identify abnormal conditions of photovoltaic equipment;

[0015] Through multi-angle background image aggregation analysis, the defect location is automatically calibrated using image coding information.

[0016] As a preferred solution of the photovoltaic power station intelligent full-scene inspection method of the present invention, the non-intrusive fault voiceprint recognition monitors whether there is a fault in the photovoltaic area equipment by analyzing the sound signal generated when the equipment is running to identify potential faults without contacting the equipment; including:

[0017] By collecting and analyzing the sound data of photovoltaic zone equipment in normal operation and fault conditions, a voiceprint database is established;

[0018] Use signal processing technology to extract the spectral characteristics, time domain characteristics and statistical characteristics of sound signals;

[0019] The extracted voiceprint features are trained and classified using deep learning algorithms to achieve automatic identification of device status.

[0020] As a preferred solution of the photovoltaic power station intelligent full-scene inspection method of the present invention, wherein: the monitoring of photovoltaic area equipment by visual parameter acquisition and processing technology includes:

[0021] Capturing images of photovoltaic area equipment through a camera;

[0022] Use image processing technology to extract visual parameters of photovoltaic area equipment failure;

[0023] Using convolutional neural networks, the model is trained to recognize and classify visual parameters, and automatically identify and locate defects and anomalies in electrical equipment.

[0024] As a preferred solution of the intelligent full-scene inspection method of photovoltaic power stations described in the present invention, the automatic transcription and identification of photovoltaic area instruments refers to the use of cameras and visual technology to identify the readings and states of instruments, and the application of convolutional neural networks for parameter extraction and classification.

[0025] The present invention also proposes a multi-agent robot that adopts the above-mentioned photovoltaic power station intelligent full-scene inspection method; it includes an energy unit for providing energy; a driving unit for driving the multi-agent robot; a collection unit for collecting information; a processing unit for processing the information collected by the collection unit; and a communication unit for transmitting and receiving networked information.

[0026] As a preferred solution of the multi-agent robot described in the present invention, the energy unit includes an energy storage module and a charging module; the drive unit includes a drive motor module, a wheel drive module, and a track drive module.

[0027] As a preferred solution of the multi-agent robot described in the present invention, the acquisition unit includes a laser radar module, an ultrasonic radar module, and a camera.

[0028] As a preferred solution of the multi-agent robot described in the present invention, the processing unit includes a graphics processor and an AI computing chip.

[0029] As a preferred solution of the multi-agent robot described in the present invention, the communication unit includes a GPS positioning module, a Beidou positioning module, a WLAN communication module, and a mobile data communication module.

[0030] The beneficial effects of the present invention are as follows: through computer vision multi-scale analysis, non-invasive fault voiceprint recognition, machine learning algorithm optimization and other means to study the intelligent diagnosis and fault feature modeling of photovoltaic power generation units and power transmission and transformation equipment within the station; through multi-bit image calibration and satellite navigation technology, accurate positioning information of equipment failures is obtained; autonomous inspection and multi-machine collaborative operation research in complex environments is carried out, and through field test verification, key technological breakthroughs in intelligent inspection of photovoltaic power stations are finally obtained, providing technical support and practical cases for unmanned / few-manned operation and maintenance of photovoltaic power stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0032] Figure 1 It is a schematic diagram of the overall structure of the multi-agent robot of the present invention. DETAILED DESCRIPTION

[0033] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0034] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0035] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0036] Secondly, the present invention is described in detail with reference to the schematic diagram. When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0037] Embodiment 1 provides a photovoltaic power station intelligent full-scene inspection method, including:

[0038] S1. Monitor the status of photovoltaic area equipment through spectral imaging technology;

[0039] Spectral imaging technology refers to the use of spectral imaging technology to provide a basis for judging the health status of photovoltaic equipment; including;

[0040] The full-spectrum imaging analysis algorithm refers to the use of spectral imaging technology to provide a basis for judging the health status of photovoltaic equipment. It collects image data under various lighting and climatic conditions, establishes a spectral feature database for photovoltaic modules and their ancillary equipment through an abnormal feature recognition algorithm, and uses annotated feature image training models based on machine learning and deep learning methods to identify abnormal conditions such as hidden cracks, hot spots, dirt, weed obstruction, foundation settlement, loose bracket fasteners, etc. on the surface of the modules. Through multi-angle background image aggregation analysis, the image coding information is used to automatically calibrate the defect location. For the images of the convergence inverter boost equipment, special annotations are made for defects such as loose cable connections, insulation aging, corrosion and overheating.

[0041] S2. Monitor whether there are faults in photovoltaic area equipment through non-invasive fault soundprint recognition;

[0042] Non-invasive fault voiceprint recognition refers to identifying potential faults by analyzing the sound signals generated by the device during operation without contacting the device. It collects and analyzes the sound data of different devices in normal operation and fault conditions to establish a voiceprint database, and then uses signal processing technology to extract the spectral characteristics, time domain characteristics and statistical characteristics of the sound signal. Further, deep learning algorithms are used to train and classify the extracted voiceprint features to achieve automatic identification of the device status.

[0043] S3, monitor photovoltaic area equipment through visual parameter acquisition and processing technology;

[0044] The machine vision parameter acquisition and processing algorithm refers to capturing the equipment image through the camera 303, and using image processing technology to extract key visual parameters such as deformation, corrosion, oil leakage and local overheating, and then using the convolutional neural network to train the model to recognize and classify these visual parameters, and automatically identify and locate the defects and abnormalities of electrical equipment.

[0045] S4. Identify photovoltaic area instruments through automated transcription.

[0046] Automated transcription is software developed for automated transcription of indoor equipment parameters, which refers to the use of high-resolution cameras 303 and machine vision technology to identify instrument readings and states, and the application of convolutional neural networks for parameter extraction and classification.

[0047] Example 2, reference Figure 1 This embodiment is different from the first embodiment in that: this embodiment also proposes a multi-agent robot, the above-mentioned photovoltaic power station intelligent full-scene inspection method; including, an energy unit 100, used to provide energy; a driving unit 200, used to drive the multi-agent robot; a collection unit 300, used to collect information; a processing unit 400, used to process the information collected by the collection unit 300; a communication unit 500, used to transmit and receive networked information.

[0048] The energy unit 100 includes an energy storage module 101 and a charging module 102; the drive unit 200 includes a drive motor module 201, a wheel drive module 202, and a track drive module 203; the acquisition unit 300 includes a laser radar module 301, an ultrasonic radar module 302, and a camera 303; the processing unit 400 includes a graphics processor 401 and an AI computing chip 402; the communication unit 500 includes a GPS positioning module 501, a Beidou positioning module 502, a WLAN communication module 503, and a mobile data communication module 504.

[0049] Among them, the energy unit 100 uses the electric energy storage module 101, that is, the battery, as the energy source, and can be replenished through the charging module 102. The charging module 102 can use fast charging technology to achieve the effect of fast energy replenishment. The drive unit 200 includes a motor module and a track drive module 203 and a wheel drive module 202, wherein the track drive module 203 and the wheel drive module 202 are respectively installed in outdoor and indoor scenes.

[0050] The acquisition unit 300 collects environmental distance information through the laser radar module 301 and the ultrasonic radar module 302, collects image information using the camera 303, establishes a three-dimensional map model through the high-performance graphics processor 401, and identifies abnormal conditions in the three-dimensional model through the AI ​​computing chip 402.

[0051] The communication unit 500 has positioning and networking functions. The real-time position of the robot is jointly determined by the GPS positioning module 501 and the Beidou positioning module 502. The 4G / 5G mobile data communication module 504 and the WAN communication module are suitable for indoor and outdoor environments.

[0052] Through the above units, the multi-agent robot can adopt autonomous navigation and obstacle avoidance algorithms. The robot's autonomous navigation and obstacle avoidance algorithm means that by carrying a laser radar module 301, a camera 303, and an ultrasonic radar module 302, the robot can build an environmental map in real time and locate its own position, thereby achieving effective cruising of the photovoltaic array. The laser radar and ultrasonic sensors are used to accurately measure the distance of the surrounding environment and generate a high-resolution three-dimensional map. The camera 303 is used to capture visual information and assist the laser radar module 301 in environmental recognition and obstacle detection. The SLAM algorithm is used in combination with sensor data to achieve real-time positioning and map construction of the robot, supporting stable cruising in complex environments.

[0053] The multi-agent robot can adopt a multi-machine collaborative inspection scheduling model, which means dynamically optimizing the robot's work deployment and inspection trajectory according to established task plans and unplanned temporary tasks, taking into account endurance, communication and operating efficiency, to achieve efficient matching of resources and tasks.

[0054] It can also accurately locate faults of massive equipment in photovoltaic areas by extracting feature points and corresponding ground control points, combining image calibration and GNSS positioning to achieve precise calibration of robot detection fault points.

[0055] The rest of the structure is the same as that of Example 1.

[0056] Importantly, it should be noted that the construction and arrangement of the present application shown in a plurality of different exemplary embodiments are only exemplary. Although only a few embodiments are described in detail in this disclosure, it should be readily understood by those who refer to this disclosure that many modifications are possible (e.g., the size, scale, structure, shape and proportion of various elements, and parameter values ​​such as temperature, pressure, etc., installation arrangements, use of materials, color, directional changes, etc.) without substantially departing from the novel teachings and advantages of the subject matter described in the application. For example, the element shown as integrally formed can be composed of multiple parts or elements, the position of the element can be inverted or otherwise changed, and the nature or number or position of the discrete element can be changed or changed. Therefore, all such modifications are intended to be included in the scope of the present invention. The order or sequence of any process or method steps can be changed or reordered according to alternative embodiments. In the claims, any "device plus function" clause is intended to cover the structure of performing the function described herein, and is not only structurally equivalent but also equivalent structure. Without departing from the scope of the present invention, other replacements, modifications, changes and omissions can be made in the design, operating conditions and arrangement of the exemplary embodiments. Therefore, the invention is not limited to a specific embodiment, but extends to numerous modifications still falling within the scope of the appended claims.

[0057] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0058] It will be appreciated that in the development of any actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but will be a routine task of design, fabrication, and production for those of ordinary skill having the benefit of this disclosure without undue experimentation.

[0059] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A photovoltaic power station intelligent full-scene inspection method, characterized by: include, Monitor the status of photovoltaic area equipment through spectral imaging technology; Monitor whether there are faults in photovoltaic area equipment through non-intrusive fault soundprint recognition; Monitoring photovoltaic area equipment through visual parameter acquisition and processing technology; and, Identify photovoltaic area instrumentation through automated transcription.

2. The photovoltaic power station intelligent full-scene inspection method according to claim 1, characterized in that: The monitoring of the status of photovoltaic area equipment by spectral imaging technology refers to using spectral imaging technology to provide a basis for judging the health status of photovoltaic area equipment; including; Collecting image data; Establish a spectral feature database of photovoltaic area equipment through an abnormal feature recognition algorithm; Based on machine learning and deep learning algorithms, the model is trained using labeled feature images to identify abnormal conditions of photovoltaic equipment; Through multi-angle background image aggregation analysis, the defect location is automatically calibrated using image coding information.

3. The photovoltaic power station intelligent full-scene inspection method according to claim 1 or 2, characterized in that: The non-intrusive fault soundprint recognition method for monitoring whether there is a fault in the photovoltaic area equipment is to identify potential faults by analyzing the sound signals generated when the equipment is running without contacting the equipment; include, By collecting and analyzing the sound data of photovoltaic zone equipment in normal operation and fault conditions, a voiceprint database is established; Use signal processing technology to extract the spectral characteristics, time domain characteristics and statistical characteristics of sound signals; The extracted voiceprint features are trained and classified using deep learning algorithms to achieve automatic identification of device status.

4. The photovoltaic power station intelligent full-scene inspection method according to claim 3, characterized in that: The monitoring of photovoltaic area equipment by visual parameter acquisition and processing technology includes: Capturing a photovoltaic area device image by means of a camera (303); Use image processing technology to extract visual parameters of photovoltaic area equipment failure; Using convolutional neural networks, the model is trained to recognize and classify visual parameters, and automatically identify and locate defects and anomalies in electrical equipment.

5. The photovoltaic power station intelligent full-scene inspection method according to claim 1 or 4, characterized in that: The identification of photovoltaic area instruments and meters by automated transcription refers to the use of a camera (303) and visual technology to identify the readings and states of the instruments and meters, and the application of a convolutional neural network to perform parameter extraction and classification.

6. A multi-agent robot, characterized in that: Adopting the photovoltaic power station intelligent full-scene inspection method described in any one of claims 1 to 5; include, An energy unit (100), used for providing energy; A driving unit (200), used for driving the multi-agent robot; A collection unit (300), used for collecting information; A processing unit (400), used for processing the information collected by the collection unit (300); The communication unit (500) is used to transmit and receive network information.

7. The multi-agent robot according to claim 6, characterized in that: The energy unit (100) comprises an energy storage module (101) and a charging module (102); The driving unit (200) comprises a driving motor module (201), a wheel driving module (202), and a track driving module (203).

8. The multi-agent robot according to claim 6 or 7, characterized in that: The acquisition unit (300) comprises a laser radar module (301), an ultrasonic radar module (302), and a camera (303).

9. The multi-agent robot according to claim 8, characterized in that: The processing unit (400) includes a graphics processor (401) and an AI computing chip (402).

10. The multi-agent robot according to claim 9, characterized in that: The communication unit (500) comprises a GPS positioning module (501), a Beidou positioning module (502), a WLAN communication module (503), and a mobile data communication module (504).

Images