Bird repelling system for photovoltaic module based on big data and autonomous deep learning

Abstract

The invention relates to the technical field of bird repelling systems, in particular to a bird repelling system based on big data and autonomous deep learning and used for a photovoltaic module, bird collecting modules and bird repelling units are distributed on the periphery of the photovoltaic module, the bird collecting modules are used for collecting the types and the number of birds close to the photovoltaic module, and the bird repelling units are used for repelling the birds. The bird repelling unit correspondingly executes a bird repelling strategy according to bird types collected by the bird collection module, the feedback module feeds back the bird repelling unit according to bird types and bird number changes before and after the bird repelling strategy is verified, then the bird repelling unit is controlled to correspondingly execute the bird repelling strategy, and the bird repelling strategy is recorded through the autonomous learning module. According to the self-adaptive bird repelling method, the type of repelled birds and the number of the repelled birds are known, a bird repelling strategy executed by the bird repelling unit is bound with the type of the repelled birds, and a perfect data model is continuously constructed, so that self-adaptive bird repelling is realized, and the cleaning difficulty and the operation and maintenance cost of the photovoltaic module are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of bird repellent systems, and more specifically, to a bird repellent system for photovoltaic components based on big data and autonomous deep learning. Background Art

[0002] Photovoltaic power generation is an efficient technology that uses the photovoltaic effect of semiconductor interfaces to directly convert light energy into electrical energy. The system mainly consists of three key parts: solar panels, controllers and inverters. Solar panels, also known as photovoltaic modules, are usually large-area modules formed by connecting and packaging multiple battery cells in series, which can effectively convert sunlight into electrical energy.

[0003] Since photovoltaic modules are exposed to the outdoor environment for a long time, they are easily attached with dust, leaves and other debris, which will directly affect the intensity of light received by them, thereby reducing the efficiency of light energy conversion. Therefore, it is particularly important to clean and maintain photovoltaic modules regularly.

[0004] In addition, birds may leave droppings when flying over, which not only further blocks light, but also becomes extremely difficult to clean due to its composition and stickiness. If bird droppings are attached to photovoltaic modules for a long time, it will increase the difficulty of maintenance and may cause damage to the battery modules.

[0005] To ensure the smoothness and power generation efficiency of photovoltaic modules, they are usually cleaned regularly in dusty areas or areas with frequent bird activities, but this undoubtedly further increases the cost of operation and maintenance. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the object of the present invention is to provide a bird repellent system based on big data and autonomous deep learning for photovoltaic components, which has the advantages of targeted expulsion of bird flocks, reducing the difficulty of cleaning photovoltaic components and reducing the operation and maintenance costs.

[0007] The above technical objectives of the present invention are achieved through the following technical solutions: A bird-repelling system based on big data and autonomous deep learning for photovoltaic modules, comprising: Bird collection module, collects the species and number of birds close to the photovoltaic modules; The bird-repelling unit executes the bird-repelling strategy according to the bird species collected by the bird collection module; The feedback module provides feedback to the bird-repelling unit based on the changes in bird species and bird numbers before and after the bird-repelling strategy is implemented; The autonomous learning module collects the bird species and bird numbers collected by the bird collection module and the non-expelled bird species and the non-expelled bird numbers after the bird-repelling unit is started by the feedback module. The expelled bird species and the expelled bird numbers are known, and the bird-repelling strategy executed by the bird-repelling unit is bound to the expelled bird species.

[0008] Preferably, the bird collection module includes a camera module and an infrared sensing component. The infrared sensing component senses the direction of the bird flock. After determining the direction of the bird flock, the camera module cooperates to capture and photograph the bird flock. According to the multiple camera photos captured by the camera module, the identification target in the camera photos is randomly framed, the color, head and body of the identification target are used as identification feature elements, and the bird species in the bird flock are identified based on the comparison between the standard database and the identification feature elements.

[0009] Preferably, a bird feature scoring model is established based on the color, head, and body identification feature elements of the identification target as weights: T = (C*W1) + (H*W2) + (B*W3); Where C: color score, W1: color weight, H: head score; W2: head weight; B: body score; WB: body weight; T: bird feature score; Based on the bird feature score, the identified bird species are sorted by bird priority, and the bird repelling unit issues bird repelling strategies in sequence for the identified bird species according to the bird priority.

[0010] Preferably, the bird collection module collects camera photos before and after the operation of the bird repelling unit for comparison, and based on the comparison between the bird species and the number of birds in the bird flock before and after the operation of the bird repelling unit, further feeds back to the bird collection module through the feedback module, and the bird collection module further analyzes the collected identification targets to identify the bird species in the bird flock, and based on the determination of the bird species in the bird flock, the bird repelling unit executes the repelling strategy on the bird flock, and the operation is cyclical until the bird flock falls off the induction of the bird collection module.

[0011] Preferably, the bird-repelling strategies implemented by the bird-repelling unit include sound wave repelling, visual repelling and odor repelling; The sound wave driving method is to play sound waves of a specific frequency to interfere with or induce the birds to stay away; The visual driving method is to scare the flock of birds by arranging flashing lights or moving visual devices; The odor driving sprays chemical odors in the direction of the bird flock to drive the bird flock away.

[0012] Preferably, the bird repellent strategies of sound wave repelling, visual repelling and odor repelling are sorted according to priority interference levels, and the corresponding bird repellent strategy is selected according to the moving direction of the bird flock collected by the bird collection module and the distance close to the photovoltaic component, first by sound wave repelling, then by adding visual repelling, and finally by adding odor repelling.

[0013] Preferably, the autonomous learning module conducts a comprehensive evaluation based on the changes in the number of birds before and after a single operation of the bird repellent unit, and records the correspondence between bird species and bird repellent strategies, and establishes a corresponding bird repellent strategy for the same bird species next time based on the comprehensive evaluation data.

[0014] Preferably, an evaluation threshold of the comprehensive evaluation is set in the autonomous learning module. When the calculated value based on the comprehensive evaluation is greater than or equal to the evaluation threshold, the autonomous learning module will feedback to the bird-repelling unit to preferentially execute the same corresponding bird-repelling strategy for the same bird species. When the calculated value based on the model of the comprehensive evaluation is less than the evaluation threshold, it will be recorded and saved.

[0015] Preferably, the autonomous learning module includes an established original large database, in which each bird species is matched with a corresponding bird repellent strategy, and a basic threshold of the bird repellent effect is set for each bird repellent strategy. If the comprehensive evaluation result is less than the basic threshold, it is judged that the bird species identified by the bird collection module may be wrong, until the corresponding bird species is found, and the recognition accuracy of the bird collection module in identifying the bird species is reversely improved.

[0016] In summary, the present invention has the beneficial effects that bird collection modules and bird repellent units are distributed around photovoltaic modules, the bird collection modules are used to collect bird species and bird numbers close to photovoltaic modules, the bird repellent unit performs corresponding bird repellent strategies according to the bird species collected by the bird collection modules, the feedback module gives feedback to the bird repellent unit according to the changes in bird species and bird numbers before and after the verification of the bird repellent strategy execution, thereby controlling the bird repellent unit to execute the bird repellent strategy in a corresponding manner, based on the comparison of bird species and bird numbers before and after the bird repellent unit executes the bird repellent strategy collected by the bird collection module, the self-learning module records the known bird species and bird numbers to be expelled, the bird repellent strategy executed by the bird repellent unit is bound to the bird species to be expelled, and the data model is continuously constructed and improved to achieve adaptive bird expulsion, thereby reducing the cleaning difficulty and operation and maintenance cost of photovoltaic modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a simplified connection diagram of an embodiment of the present invention; Figure 2 Schematic diagram of the division of the bird-repelling strategy according to an embodiment of the present invention.

[0018] Figure numerals: 1. Bird collection module; 11. Bird repelling unit; 12. Feedback module; 13. Autonomous learning module; 14. Original large database; 2. Sound wave repelling; 21. Visual repelling; 22. Odor repelling; 3. Photovoltaic module. DETAILED DESCRIPTION

[0019] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] It should be noted that when a component is referred to as being "fixed to" or "disposed on" another component, it can be directly on the other component or indirectly on the other component. When a component is referred to as being "connected to" another component, it can be directly or indirectly connected to the other component.

[0021] It should be understood that the orientation or position relationship indicated by terms such as "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0022] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0023] A bird repellent system for photovoltaic modules based on big data and autonomous deep learning, see Figure 1 ,include: A bird collection module 1 collects the species and number of birds near the photovoltaic module 3; The bird-repelling unit 11 executes a corresponding bird-repelling strategy according to the bird species collected by the bird collection module 1; The feedback module 12 provides feedback to the bird-repelling unit 11 according to the changes in bird species and bird numbers before and after the bird-repelling strategy is verified; The autonomous learning module 13 collects the bird species and the number of birds collected by the bird collection module 1 and the non-expelled bird species and the number of birds collected by the feedback module 12 after starting the bird-repelling unit 11. The expelled bird species and the number of birds are known, and the bird-repelling strategy executed by the bird-repelling unit 11 is bound to the expelled bird species.

[0024] In this embodiment, a bird collection module 1 and a bird repelling unit 11 are distributed around the photovoltaic component 3. The bird collection module 1 collects the bird species and the number of birds close to the photovoltaic component 3, and the bird repelling unit 11 performs a corresponding bird repelling strategy according to the bird species collected by the bird collection module 1. The feedback module 12 provides feedback to the bird repelling unit 11 according to the changes in the bird species and the number of birds before and after the bird repelling strategy is verified, thereby controlling the bird repelling unit 11 to perform the bird repelling strategy in correspondence. Based on the comparison of the bird species and the number of birds collected by the bird collection module 1 before and after the bird repelling unit 11 executes the bird repelling strategy, the self-learning module 13 records the known bird species and the number of birds to be expelled, and the bird repelling strategy executed by the bird repelling unit 11 is bound to the bird species to be expelled, and the data model is continuously constructed and improved to achieve adaptive bird expulsion, thereby reducing the cleaning difficulty and operation and maintenance cost of the photovoltaic component 3.

[0025] The bird collection module 1 includes sensors such as a camera module, an infrared sensing component, and a radar. The infrared sensing component or the radar senses the direction of the bird flock. After determining the direction of the bird flock, the camera module is used to capture and film the bird flock. Based on multiple camera photos captured by the camera module, the identification target in the camera photos is randomly framed, and the color, head, body, wings, etc. of the identification target are used as identification feature elements. Based on the comparison between the standard database and the identification feature elements, the bird species in the flock are identified.

[0026] Due to the complexity of bird species and the fact that bird flocks are in a mobile state, errors often occur when identifying bird species in a flock. Different bird species are adapted to different bird-repelling strategies. Therefore, a bird feature scoring model is established based on the color, head, and body identification feature elements of the target as weights: T = (C*W1) + (H*W2) + (B*W3); Where C: color score, W1: color weight, H: head score; W2: head weight; B: body score; WB: body weight; T: bird feature score; Based on the bird feature score, the identified bird species are sorted by bird priority, and the bird repelling unit 11 executes the bird repelling strategy for the identified bird species according to the bird priority.

[0027] The proportion of weights is obtained through a large amount of data and experiments, which will not be elaborated here one by one.

[0028] In the case of known weights, the bird species collected by the recognition module are identified, and the bird feature score of the identification target corresponding to the various bird species is calculated according to the familiarity score of the various birds in the color, head and body of the identification target. The corresponding bird species are sorted by bird priority. According to the bird priority sorting, the bird repellent unit 11 issues bird repellent strategies for the identified bird species in turn.

[0029] After the bird-repelling unit 11 executes a bird-repelling strategy each time, the bird collection module 1 collects and compares the camera photos before and after the operation of the bird-repelling unit 11. Based on the comparison between the bird species and the number of birds in the bird flock before and after the operation of the bird-repelling unit 11, the feedback module 12 further feeds back to the bird collection module 1. The bird collection module 1 further analyzes the collected identification targets to identify the bird species in the bird flock. Based on the determination of the bird species in the bird flock, the bird-repelling unit 11 executes the driving strategy on the bird flock, and the operation is cyclical until the bird flock falls out of the sensing range of the bird collection module 1.

[0030] In addition, with respect to the bird-repelling strategy executed by the bird-repelling unit 11, the degree and manner of disturbing the birds in the flock are different due to the different species of the birds.

[0031] The bird repelling strategies of this embodiment include sound wave repelling 2, visual repelling 21 and odor repelling 22.

[0032] See also Figure 2 , wherein the sound wave driving 2 plays sound waves of a specific frequency to interfere with or induce the birds to stay away; The visual driving 21 disturbs the flock of birds by arranging flashing lights or moving visual devices; The odor repellent 22 sprays chemical odors toward the direction of the bird flock to repel the bird flock.

[0033] Since visual repelling 21 is affected by the propagation distance and propagation area, odor repelling 22 is affected by the external environment due to the duration, and odor repelling 22 also needs to be supplemented with chemical reagents. For this reason, this application combines the advantages and disadvantages of the three bird repelling strategies, and combines sound wave repelling 2, visual repelling 21 and odor repelling 22 for application.

[0034] Specifically, the bird repelling strategies of sound wave repelling 2, visual repelling 21 and odor repelling 22 are sorted by priority interference level, and the corresponding bird repelling strategy is selected according to the distance of the moving direction of the bird flock collected by the bird collection module 1 close to the photovoltaic component 3, first through the sound wave repelling 2, then by adding visual repelling 21, and finally adding odor repelling 22.

[0035] Bird repelling strategies also include sound repelling at specific frequencies2; visual repelling with different light colors21, and odor repelling with different smells22.

[0036] The bird collection module 1 collects the moving direction of the bird flock and the distance from the photovoltaic module 3, and adopts different bird repelling strategies in sequence. The first is the sonic repelling 2, which has a large diffusion area and is accurate. Secondly, the visual repelling 21 is used for interference, and finally the smell repelling 22 is used to further enhance the repelling effect.

[0037] The bird repelling strategies of sound repelling 2, visual repelling 21 and odor repelling 22 are used to prioritize the interference levels, which can avoid wasting resources and expel and disturb the bird species to the maximum extent.

[0038] In different areas, three different bird-repelling strategies, namely, sound repelling 2, visual repelling 21 and odor repelling 22, are established to avoid mutual influence between different bird-repelling strategies and prevent birds from flying randomly.

[0039] Secondly, the autonomous learning module 13 described in this application conducts a comprehensive evaluation based on the changes in the number of birds before and after a single operation of the bird-repelling unit 11, and records the correspondence between bird species and bird-repelling strategies, and establishes a corresponding bird-repelling strategy for the same bird next time based on the comprehensive evaluation data.

[0040] In actual operation, after the bird-repelling strategy is implemented, a comprehensive evaluation is conducted based on the changes in the number of birds before and after, to determine which sound wave corresponds to which bird species with the most obvious repelling effect. The corresponding relationship between the bird species and the sound wave is bound and recorded, and can be used as a bird-repelling strategy for the same bird species next time.

[0041] Since different bird-repelling strategies have different effects on the expulsion of the same bird species, the autonomous learning module 13 can be used to find the best bird-repelling strategy for the same bird species among the different bird-repelling strategies that can be executed by the bird-repelling unit 11.

[0042] Based on this, by setting an evaluation threshold of the comprehensive evaluation in the autonomous learning module 13, when the calculated value based on the comprehensive evaluation is greater than or equal to the evaluation threshold, the autonomous learning module 13 will feedback to the bird-repelling unit 11 to preferentially execute the same corresponding bird-repelling strategy for the same bird species; when the calculated value based on the comprehensive evaluation model is less than the evaluation threshold, it will be recorded and saved.

[0043] In addition, an original big database 14 is established in the autonomous learning module 13. A corresponding bird-repelling strategy is matched for each bird species in the original big database 14. A basic threshold of the bird-repelling effect is set for each bird-repelling strategy. If the comprehensive evaluation result is less than the basic threshold, it is judged that the bird species identified by the bird collection module 1 may be wrong, until the corresponding bird species is found, thereby reversely improving the recognition accuracy of the bird collection module 1 in identifying the bird species.

[0044] The bird-repelling effect here refers to the change in the number of birds before and after the bird-repelling unit 11 is executed. If the comprehensive evaluation result is less than the basic threshold, it is judged that the bird species identified by the bird collection module 1 may be wrong. Based on the acquisition of the bird species in the identified bird flock by the bird collection module 1, the birds in the bird priority sorting are ranked according to the identified bird species, and the bird-repelling strategy for the next level of bird species is corresponding, until the bird flock is driven away, and the recognition accuracy of the bird collection module 1 in identifying the bird species is improved.

[0045] The above embodiments are merely explanations of the present invention and are not limitations of the present invention. After reading this specification, those skilled in the art may make modifications to the embodiments without any creative contribution as needed. However, such modifications are protected by the patent law as long as they are within the scope of the claims of the present invention.

Claims

1. A bird repelling system for photovoltaic modules based on big data and autonomous deep learning, characterized by: include: Bird collection module, collects the species and number of birds close to the photovoltaic modules; The bird-repelling unit executes the bird-repelling strategy according to the bird species collected by the bird collection module; The feedback module provides feedback to the bird-repelling unit based on the changes in bird species and bird numbers before and after the bird-repelling strategy is implemented; The autonomous learning module collects the bird species and bird numbers collected by the bird collection module and the non-expelled bird species and the non-expelled bird numbers after the bird-repelling unit is started by the feedback module. The expelled bird species and the expelled bird numbers are known, and the bird-repelling strategy executed by the bird-repelling unit is bound to the expelled bird species.

2. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 1 is characterized in that: The bird collection module includes a camera module and an infrared sensing component. The infrared sensing component senses the direction of the bird flock. After determining the direction of the bird flock, the camera module cooperates to capture and film the bird flock. According to the multiple camera photos captured by the camera module, the identification target in the camera photos is randomly framed, the color, head and body of the identification target are used as identification feature elements, and the bird species in the bird flock are identified based on the comparison between the standard database and the identification feature elements.

3. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 2 is characterized in that: A bird feature scoring model is established based on the identification feature elements of the target’s color, head, and body as weights: T = (C*W1) + (H*W2) + (B*W3); in C: color score, W1: color weight, H: head score; W2: head weight; B: Body score; WB: body weight; T: bird feature scoring score; Based on the bird feature score, the identified bird species are sorted by bird priority, and the bird repelling unit issues bird repelling strategies in sequence for the identified bird species according to the bird priority.

4. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 3 is characterized by: The bird collection module collects and compares the camera photos before and after the operation of the bird repelling unit. Based on the comparison between the bird species and the number of birds in the bird flock before and after the operation of the bird repelling unit, the feedback is further fed back to the bird collection module through the feedback module. The bird collection module further analyzes the collected identification targets to identify the bird species in the bird flock. Based on the determination of the bird species in the bird flock, the bird repelling unit executes the repelling strategy for the bird flock, and the operation is cyclical until the bird flock falls off the induction of the bird collection module.

5. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 4 is characterized in that: The bird-repelling strategies implemented by the bird-repelling unit include acoustic repelling, visual repelling and odor repelling; The sound wave driving method is to play sound waves of a specific frequency to interfere with or induce the birds to stay away; The visual driving method is to scare the flock of birds by arranging flashing lights or moving visual devices; The odor driving sprays chemical odors in the direction of the bird flock to drive the bird flock away.

6. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 5 is characterized by: The bird repellent strategies of sound wave repellent, visual repellent and odor repellent are sorted by priority interference level. According to the moving direction of the bird flock collected by the bird collection module and the distance close to the photovoltaic component, the corresponding bird repellent strategy is selected, first by sound wave repellent, then by adding visual repellent, and finally by adding odor repellent.

7. The bird repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 1 is characterized in that: The autonomous learning module performs a comprehensive evaluation based on the changes in the number of birds before and after a single operation of the bird-repelling unit, and records the corresponding relationship between the bird species and the bird-repelling strategy, and establishes a corresponding bird-repelling strategy for the same bird species next time based on the comprehensive evaluation data.

8. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 7 is characterized in that: An evaluation threshold for comprehensive evaluation is set in the autonomous learning module. When the calculated value based on the comprehensive evaluation is greater than or equal to the evaluation threshold, the autonomous learning module will feedback to the bird-repelling unit to preferentially execute the same corresponding bird-repelling strategy for the same bird species. When the calculated value based on the model of comprehensive evaluation is less than the evaluation threshold, it will be recorded and saved.

9. The bird-repelling system based on big data and autonomous deep learning for photovoltaic modules according to claim 7, characterized in that: The autonomous learning module includes an established original large database, in which each bird species is matched with a corresponding bird repelling strategy, and a basic threshold of the bird repelling effect is set for each bird repelling strategy. If the comprehensive evaluation result is less than the basic threshold, it is judged that the bird species identified by the bird collection module may be wrong, until the corresponding bird species is found, and the recognition accuracy of the bird collection module in identifying the bird species is reversely improved.

Images