A base station direct current comprehensive management and control device

Abstract

The present application relates to a kind of base station direct current comprehensive management and control device, belong to communication base station direct current power supply technical field.The device includes switching power supply, metering control module, direct current magnetic latching contactor, schottky diode, main control module, 4G communication module, visual identification module, base station standby battery, OLT standby battery, camera, infrared sensor, ultrasonic sensor, mains detection module, bluetooth module and the structure of FSU.The present application not only can realize the comprehensive management and control of base station direct current, can be carried out direct current management and control to single operator direct current load, also can realize the real-time monitoring of remote mountainous area and urban public security monitoring blind spot area communication base station, easy to popularize and apply.

Description

Technical Field

[0001] This invention belongs to the field of DC power supply technology for communication base stations, and specifically relates to a comprehensive DC control device for base stations. Background Technology

[0002] A base station, or public mobile communication base station, is an interface device for mobile devices to access the Internet. It is also a form of radio station, referring to a radio transceiver station that transmits and receives information between a mobile communication switching center and a mobile phone terminal within a certain radio coverage area. The construction of mobile communication base stations is a significant investment for mobile communication operators, and its construction is generally based on factors such as coverage, call quality, investment efficiency, ease of construction, and convenience of maintenance. As mobile communication network services develop towards data-driven and packet-based communication, the development trend of mobile communication base stations will inevitably be towards broadband, large-scale coverage, and IP-based deployment.

[0003] The following problems exist in the current integrated DC management and control of base stations:

[0004] 1. Inaccuracies in power management

[0005] (1) Limitations of a single copper busbar design

[0006] Current base station systems typically use parallel copper busbars to expand DC output ports. While this design simplifies the hardware structure, it lacks independent control and monitoring capabilities for each output port. This deficiency is particularly pronounced in shared base station environments, where the power consumption of different operators' equipment cannot be monitored individually. This makes it impossible to accurately calculate the actual power consumption of each user, forcing the allocation of power costs through average distribution or estimation. This inaccuracy can easily lead to disputes between operators and increases management complexity.

[0007] (2) Lack of digital load management

[0008] Traditional base station power supply systems lack real-time voltage, current, and power monitoring capabilities for each branch load, making it impossible to achieve refined management of each electrical device. This results in the system's inability to detect and respond promptly when an anomaly occurs in a load (such as overcurrent or undervoltage), increasing the risk of power equipment failure. Furthermore, the shortcomings of traditional systems are also reflected in their methods of power data acquisition and storage, lacking intelligent data analysis capabilities.

[0009] 2. The single nature and low reliability of the backup power system

[0010] (1) Limitations of single-cell backup power

[0011] Currently, most communication base stations use multiple battery banks as backup power for all DC loads. While this design can handle short-term mains power outages, operators prefer to use individual batteries for critical loads such as base station OLT equipment due to the importance of backup power levels. Furthermore, in multi-user base station environments, a single-battery system cannot provide independent power support for different users' equipment.

[0012] (2) Lack of intelligent switching and energy scheduling

[0013] Traditional base station backup power systems lack intelligent power dispatching and switching capabilities. Typically, when mains power is interrupted, the system simply switches to battery power without considering the priorities of different devices or the needs of different loads. This indiscriminate switching method easily leads to rapid battery depletion, failing to guarantee priority power supply to critical equipment, and thus affecting the continuous operation of the base station.

[0014] 3. High cost and limitations of anti-theft monitoring systems

[0015] (1) Traditional camera solutions are expensive

[0016] Currently, some base stations have installed high-resolution surveillance cameras to prevent equipment theft. While these cameras can provide 24 / 7 real-time monitoring, their high cost, complex installation, and subsequent maintenance expenses make them difficult to implement in remote areas or unattended base stations. Furthermore, these cameras rely on mains power; when the mains power fails, their anti-theft function also fails, leaving the base station equipment unprotected in the event of a power outage.

[0017] (2) Insufficient identification and storage

[0018] Traditional surveillance systems typically only record video and store it locally or on a remote server, lacking real-time analysis capabilities. Cameras cannot identify or distinguish between unauthorized personnel and base station administrators, rendering the system ineffective in actual security.

[0019] Therefore, overcoming the shortcomings of existing technologies is a problem that urgently needs to be solved in this field. Summary of the Invention

[0020] The purpose of this invention is to overcome the shortcomings of the prior art and provide a base station DC integrated management and control device.

[0021] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0022] A base station DC integrated management and control device includes a switching power supply, a metering control module, a DC magnetic latching contactor, a Schottky diode, a main control module, a 4G communication module, a visual recognition module, a base station backup battery, an OLT backup battery, a camera, an infrared sensor, an ultrasonic sensor, a mains power detection module, a Bluetooth module, and an FSU.

[0023] Three-phase AC power is connected to a switching power supply that outputs 48V DC power.

[0024] The metering control module, OLT equipment, and OLT backup battery are cascaded in sequence.

[0025] The 0V terminal of the switching power supply is connected to the positive terminal of the metering control module, one end of the DC magnetic latching contactor, and the positive terminal of the Schottky diode, respectively.

[0026] The other end of the DC magnetic latching contactor is connected to the negative terminal of the Schottky diode, the positive terminal of the OLT equipment, and the positive terminal of the OLT backup battery, respectively.

[0027] The -48V terminal of the switching power supply is connected to the negative terminal of the metering control module, the negative terminal of the OLT equipment, and the negative terminal of the OLT backup battery, respectively.

[0028] The switching power supply is also connected to the base station's backup battery;

[0029] The mains power detection module is connected to the switching power supply;

[0030] The main control module is connected to the metering control module, DC magnetic latching contactor, vision recognition module, and mains power detection module, respectively.

[0031] There are multiple metering control modules; each metering control module is connected to a DC load of an operator.

[0032] The visual recognition module is connected to the camera, infrared sensor, and ultrasonic sensor respectively;

[0033] The main control module is connected to the cloud platform via a 4G communication module, to the WeChat mini-program via a Bluetooth module, and to the tower maintenance platform via the FSU.

[0034] Furthermore, the mains power detection module is used to check whether the input received by the switching power supply is mains power;

[0035] When the mains power detection module detects mains power, the main control module controls the DC magnetic latching contactor to close, and the DC power supplies the OLT equipment and charges the OLT backup battery through the DC magnetic latching contactor.

[0036] When the mains power detection module detects no mains power, the main control module controls the DC magnetic latching contactor to open, and the operator's DC load is powered by the base station's backup battery; the OLT equipment is powered by the OLT backup battery alone.

[0037] The metering control module is used to monitor the load status of the DC load of the operator connected to it in real time and transmit the monitored load status to the main control module; the load status includes voltage, current and power.

[0038] The main control module determines whether an abnormal load situation has occurred based on the received load status. If an abnormal load situation occurs, it triggers an alarm or controls the metering control module to perform protective measures.

[0039] The camera is used for image acquisition at the base station;

[0040] Infrared sensors are used for real-time acquisition of thermal images of base stations;

[0041] Ultrasonic sensors are used to collect ultrasonic data from base stations;

[0042] The camera is activated when the infrared sensor detects changes in the thermal image caused by a person or animal entering the base station, or when the ultrasonic sensor detects ultrasonic waves caused by the operation of equipment outside the base station; otherwise, the camera is in sleep mode.

[0043] The visual recognition module is used to determine whether a person or animal has entered the base station based on the data collected by the camera, infrared sensor, and ultrasonic sensor. If it is an animal, an alarm is triggered. If it is a person entering the base station, it is necessary to determine whether the person has access permission. If the person does not have access permission, an alarm is triggered. If the person has access permission, it is further determined whether the person has any abnormal behavior. If abnormal behavior is found, an alarm is triggered.

[0044] The visual recognition module sends the data collected by the camera, infrared sensor, and ultrasonic sensor, along with the results determined by the visual recognition module, to the main control module.

[0045] The main control module sends its interactions with the metering control module and the visual recognition module, as well as the status of the DC magnetic holding contactor and the detection results of the mains power detection module, to the cloud platform via the 4G communication module, to the WeChat mini-program via the Bluetooth module, and to the tower operation and maintenance platform via the FSU.

[0046] Furthermore, it also includes a lightning protection and surge absorption module, and the lightning protection and surge absorption module, metering and control module, OLT equipment, and OLT backup battery are cascaded in sequence;

[0047] The switching power supply outputs 48V DC power. After the surge protection and surge absorption module suppresses and absorbs the high-frequency pulses from the previous stage, it is supplied to the operator's DC load through the DC metering and control module.

[0048] Furthermore, the switching power supply includes an AC / DC module and a two-stage low-voltage disconnect contactor;

[0049] The three-phase AC power is connected to the AC / DC module of the switching power supply, which outputs 48V DC power. After voltage detection by the secondary low-voltage disconnect contactor of the switching power supply, a controllable output is achieved.

[0050] Furthermore, when the OLT backup battery fails or is not connected, the OLT equipment is powered by the base station backup battery.

[0051] Furthermore, in the main control module, the abnormal load conditions include overcurrent, undervoltage, and overload; when overcurrent or undervoltage occurs, an alarm is triggered; when overload occurs, the metering control module is controlled to perform protection measures; the protection measures are to disconnect the DC output and automatically restore the DC output after the abnormality is resolved.

[0052] Furthermore, the main control module is also used to calculate the electricity consumption of each operator's DC load based on the transmitted load status. Simultaneously, the main control module distinguishes the electricity consumption by time, dividing it into today's electricity consumption, yesterday's electricity consumption, this month's electricity consumption, last month's electricity consumption, and total electricity consumption for each operator, as well as the total electricity consumption of this device and yesterday's electricity consumption. The main control module multiplies each operator's today's electricity consumption, yesterday's electricity consumption, this month's electricity consumption, last month's electricity consumption, and total electricity consumption by a pre-set electricity price to obtain the electricity cost data for each operator. After storing the above data locally, the main control module sends it to the FSU and 4G communication modules via the RS485 interface. Upon receiving the data, the FSU and 4G communication modules respectively send it to the tower maintenance platform and the cloud platform. Both the tower maintenance platform and the cloud platform display the data and export reports based on the received data, serving as the original basis for electricity cost settlement.

[0053] Furthermore, the main control module processes the interaction content between the pre-defined standard wired and wireless protocol texts and the metering control module, the visual recognition module, the status of the DC magnetic latching contactor, and the detection results of the mains power detection module before sending them to the 4G communication module, Bluetooth module, and FSU. After parsing, the FSU uploads the data to the tower operation and maintenance platform through its built-in wireless module.

[0054] The 4G communication module sends data to the cloud platform;

[0055] After the Bluetooth module establishes a connection with the WeChat mini-program on the mobile phone, it sends data to the WeChat mini-program to achieve local data monitoring.

[0056] Furthermore, the other end of the DC magnetic latching contactor is also connected to the positive terminal of the camera, and the -48V terminal of the switching power supply is also connected to the negative terminal of the camera.

[0057] Furthermore, the visual recognition module incorporates built-in facial recognition and behavior analysis models;

[0058] The face recognition and behavior analysis model includes a hierarchical feature extraction module, an adaptive fusion module, a key point feature extraction module, an adaptive convolution module, a multi-level sparse coding module, and a behavior recognition and temporal processing module;

[0059] The hierarchical feature extraction module uses multi-scale pyramid decomposition and then lightweight convolution to convolve the image at each scale to obtain feature maps at different levels.

[0060] The adaptive fusion module is used to fuse feature maps at different levels to obtain a fused feature map.

[0061] The key point feature extraction module is used to detect key points in all feature maps to achieve face recognition;

[0062] The adaptive convolution module is used to adaptively adjust the convolution operation according to the complexity of the image, retaining only the most important convolution kernels;

[0063] The multi-level sparse coding module is used to obtain the sparse coding vectors of all feature maps and merge them to obtain the sparsed feature vectors at each time step.

[0064] The behavior recognition and temporal processing module is used to perform behavior recognition based on the sparse feature vectors at each time step.

[0065] When the mains power detection module detects mains power, the main control module controls the DC magnetic latching contactor to close, and the DC power supplies the OLT equipment and charges the OLT backup battery through the DC magnetic latching contactor. The DC magnetic latching contactor is introduced in this invention because its contact resistance is very low, which can effectively avoid the problem of Schottky diodes overheating due to long-term operation.

[0066] When the mains power detection module detects no mains power, the main control module controls the DC magnetic latching contactor to open, and the operator's DC load is powered by the base station's backup battery. Due to the presence of Schottky diodes, when the OLT backup battery voltage is higher than the base station backup battery voltage, the OLT equipment is powered solely by the OLT backup battery. Under normal circumstances, the power consumption of the OLT equipment is much lower than that of other DC loads in the base station, such as AAU (Active Antenna Unit) and RRU (Remote Radio Unit). Therefore, under normal circumstances, after a mains power outage, the voltage of the OLT backup battery will be higher than that of the base station backup battery.

[0067] This invention employs a dual backup power system and an independent battery management mode. Specifically, it includes two independent battery packs: a base station backup battery (providing basic backup power for all DC loads) and an OLT backup battery (dedicated to independent power supply for critical equipment, such as OLT devices). Preferably, each battery pack is equipped with a power monitoring module and a battery management system (BMS). The power monitoring module is used to monitor battery status in real time (such as voltage, current, and temperature); the battery management system is used to ensure the reliability of the battery during charging, discharging, and operation.

[0068] When the mains power fails, the device of this invention first switches to the base station backup battery, while the OLT backup battery provides independent power to the OLT equipment. When power is scarce, the device of this invention prioritizes ensuring the continuous operation of the OLT equipment, and gradually shuts down secondary loads in the operator's DC load according to the low-voltage shutdown thresholds and backup power durations set in advance for each channel of the main control module via WeChat mini-program, tower maintenance platform, or cloud platform, in order to extend the OLT system's operating time.

[0069] The device of this invention can be flexibly expanded. The number of battery packs can be increased or decreased according to the DC load requirements of the operator, or different battery capacities can be configured to adapt to the power supply requirements in different scenarios.

[0070] The metering control module of this invention is equipped with voltage, current, power monitoring sensors and temperature sensors to monitor the load status of the connected operator's DC load in real time for each DC output (used by the operator's DC load), and transmits the monitored load status to the main control module; the load status includes voltage, current, power and temperature; preferably, the metering control module is also equipped with a microcontroller (MCU), which transmits the load status to the main control module; more preferably, the metering control module communicates with the main control module via an RS485 bus.

[0071] The main control module determines whether an abnormal load condition has occurred based on the received load status. The abnormal load conditions include overcurrent, undervoltage, overload, and overheating. If any of these conditions occur, an alarm is triggered or the metering control module is controlled to perform protective measures. Specifically, when overcurrent or undervoltage occurs, an alarm is triggered; when overload or overheating occurs, the metering control module is controlled to perform protective measures. The protective measures involve disconnecting the DC output of that circuit and automatically restoring the DC output of that circuit once the abnormality is resolved.

[0072] In this invention, the main control module is also used to calculate the electricity consumption of each operator's DC load based on the transmitted load status. Simultaneously, the main control module distinguishes the electricity consumption by time, dividing it into today's electricity consumption, yesterday's electricity consumption, this month's electricity consumption, last month's electricity consumption, total electricity consumption, and the device's total electricity consumption and yesterday's electricity consumption. The main control module calculates the electricity cost data for each operator by multiplying the pre-set electricity price by the electricity consumption of each operator (today's electricity consumption, yesterday's electricity consumption, this month's electricity consumption, last month's electricity consumption, and total electricity consumption). After storing the above data locally, the main control module sends it to the FSU and 4G communication modules via the RS485 interface. Upon receiving the data, the FSU and 4G communication modules respectively send it to the tower maintenance platform and cloud platform. Both the tower maintenance platform and cloud platform display the data and export reports based on the above data, serving as the original basis for electricity cost settlement. Thus, this device can track and record the electricity consumption of each operator's DC load in real time, ensuring the fairness and accuracy of power allocation.

[0073] In this invention, the predetermined standard wired protocol text adopts RS485, but the data interaction with the visual recognition module adopts USB; the wireless protocol text adopts Modbus RTU.

[0074] In this invention, the data collected by the visual recognition module is exchanged with the main control module via an RS485 bus.

[0075] In this invention, the secondary low-voltage disconnect contactor inside the switching power supply has a voltage divider detection circuit for detecting DC voltage and a shunt for detecting DC current. After detection, controllable output can be achieved.

[0076] This invention employs a low-cost embedded AI platform (visual recognition module) equipped with a 5MP wide-angle camera to achieve local facial recognition and behavior analysis. Furthermore, it can perform real-time recognition and determine whether a person is an authorized operator without relying on a network. This design reduces dependence on cloud resources, significantly lowering latency and costs.

[0077] The visual recognition module incorporates facial recognition and behavior analysis algorithms, enabling it to detect and identify authorized personnel and their abnormal behavior, such as damaging equipment. Once abnormal behavior is detected, an alarm is triggered, and the collected images and alarm information are transmitted to the cloud platform via the 4G communication module, ensuring that management personnel are promptly aware of the situation on-site.

[0078] To meet the low-power requirements of remote base stations, this invention is designed for multi-mode operation: it enters a low-power standby mode when no one or animal is near; the camera only takes pictures when the infrared sensor detects changes in the thermal image due to someone or an animal entering the base station, or when the ultrasonic sensor detects ultrasonic waves caused by equipment operating outside the base station. At this time, the visual recognition module initiates face recognition and behavior analysis. This event-driven design reduces power consumption, significantly extends the camera's lifespan, and avoids the storage burden associated with 24 / 7 recording. Preferably, the camera's power supply is taken from the OLT backup battery side of this device. Even in the event of a mains power outage, the OLT backup battery powers the camera, ensuring its normal operation.

[0079] This invention combines facial recognition and behavior analysis algorithms to achieve local facial data collection and storage. When a new facial image is collected, it is uploaded to the cloud (i.e., cloud platform) in real time via a 4G communication module to achieve the purpose of preventing theft and evidence collection and real-time management.

[0080] The tower operation and maintenance platform, cloud platform, and mobile WeChat mini-program used in this invention are all existing technologies, and this invention does not limit them.

[0081] This invention preferably employs edge computing technology to process facial recognition and data compression locally, and ensures data security through encrypted transmission. Even if the data within the device's camera, infrared sensor, and ultrasonic sensor is maliciously tampered with or forcibly removed, the relevant data will be stored in the device's visual recognition module for subsequent investigation.

[0082] This invention extracts and fuses face and behavior features in a hierarchical manner through a hierarchical feature extraction module, introduces an adaptive convolution module to dynamically adjust computing resources, and combines multi-level sparse coding to compress and simplify feature representations, thereby achieving efficient and low-power face and behavior recognition.

[0083] Among them, the hierarchical feature extraction module is used to extract multi-scale and multi-level facial and behavioral features, which are gradually refined and fused from low to high level.

[0084] Adaptive Convolution Module (ACM): Used to dynamically adjust convolution operations, adaptively allocating computational resources based on the complexity of the input features.

[0085] Multilevel Sparse Coding Module (MSC): Used to combine multilevel sparse coding to compress and simplify feature representations to reduce computation.

[0086] This invention discloses a base station DC integrated management and control device, which solves the following technical problems:

[0087] First, it solved the problem of remote on / off management of individual DC loads within a communication base station;

[0088] Secondly, through structures such as the mains power detection module, the problem of current communication base stations taking corresponding actions for a single DC load based on the status of the external mains power or the AC generator power generation status has been solved.

[0089] Third, it solves the problem that the power supply system in existing base stations does not support digital management of the loads of multiple users;

[0090] Fourth, it solved the problem of low-cost independent backup power for transmission equipment within the base station;

[0091] Fifth, it solves the problems of traceability and theft prevention in base station operations, and realizes intelligent and digital management of DC loads of base station operators.

[0092] Compared with the prior art, the beneficial effects of this invention are as follows:

[0093] 1. By using the device of the present invention, the DC load of the operator can be intelligently controlled, and the voltage, current and power of a single DC load can be sampled and analyzed; and the DC load of a single operator can be remotely switched on and off, with active alarms for overcurrent and undervoltage, ensuring the reliable operation of the DC terminal load from the power supply side.

[0094] 2. By using the device of the present invention, the cause of abnormal power supply to a single operator's DC load can be determined, such as overcurrent, undervoltage, overload, etc. If the metering control module does not upload any alarm data, the power supply is determined to be normal.

[0095] 3. By using the device of the present invention, real-time monitoring of communication base stations in remote mountainous areas and urban monitoring blind spots can be realized. If the base station equipment is operated without the consent of the base station manager, the base station manager can know in real time through mobile APP and WEB terminal. It can realize monitoring and evidence collection for specific application scenarios at extremely low cost. Attached Figure Description

[0096] Figure 1 This is a schematic diagram of the structure of an embodiment of the base station DC integrated management and control device of the present invention; Figure 2 This is a schematic diagram of another embodiment of the base station DC integrated control device of the present invention; Figure 3 This is a schematic diagram of another embodiment of the DC integrated control device for base stations of the present invention;

[0097] The components include: 1. Switching power supply; 2. Metering control module; 3. DC magnetic latching contactor; 4. Schottky diode; 5. Main control module; 6. 4G communication module; 7. Visual recognition module; 8. Base station backup battery; 9. Operator DC load; 10. OLT equipment; 11. OLT backup battery; 12. Camera; 13. Infrared sensor; 14. Ultrasonic sensor; 15. Mains power detection module; 16. Bluetooth module; 17. FSU; 18. Lightning protection and surge absorption module; 19. AC / DC module; and 20. Secondary low-voltage disconnect contactor. Detailed Implementation

[0098] The present invention will now be described in further detail with reference to the embodiments.

[0099] Those skilled in the art will understand that the following embodiments are for illustrative purposes only and should not be construed as limiting the scope of the invention. Where specific techniques or conditions are not specified in the embodiments, they are performed in accordance with the techniques or conditions described in the literature in the field or according to the product instructions. Materials or equipment whose manufacturers are not specified are all conventional products that can be obtained by purchase.

[0100] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” to another element, it can be directly connected to the other element, or there may be an intermediate element. Furthermore, the term “connected” as used herein can include wireless connections. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0101] In the description of this invention, unless otherwise stated, "a plurality of" means two or more. The terms "inner," "upper," "lower," etc., indicate the orientation or state relationship based on the orientation or state relationship shown in the drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention.

[0102] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installed," "connected," and "equipped with" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art will understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0103] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0104] Example 1

[0105] like Figure 1 As shown, a base station DC integrated management and control device includes a switching power supply 1, a metering control module 2, a DC magnetic latching contactor 3, a Schottky diode 4, a main control module 5, a 4G communication module 6, a visual recognition module 7, a base station backup battery 8, an OLT backup battery 11, a camera 12, an infrared sensor 13, an ultrasonic sensor 14, a mains power detection module 15, a Bluetooth module 16, and an FSU 17.

[0106] Three-phase AC power is connected to switching power supply 1 to output 48V DC power;

[0107] Metering control module 2, OLT device 10, and OLT backup battery 11 are cascaded in sequence;

[0108] The 0V terminal of the switching power supply 1 is connected to the positive terminal of the metering control module 2, one end of the DC magnetic latching contactor 3, and the positive terminal of the Schottky diode 4, respectively.

[0109] The other end of the DC magnetic latching contactor 3 is connected to the negative terminal of the Schottky diode 4, the positive terminal of the OLT device 10, and the positive terminal of the OLT backup battery 11, respectively.

[0110] The -48V terminal of the switching power supply 1 is connected to the negative terminal of the metering control module 2, the negative terminal of the OLT device 10, and the negative terminal of the OLT backup battery 11, respectively.

[0111] The switching power supply 1 is also connected to the base station backup battery 8;

[0112] The mains power detection module 15 is connected to the switching power supply 1 and is used to check whether the input received by the switching power supply 1 is mains power;

[0113] When the mains power detection module 15 detects that the mains power is available, the main control module 5 controls the DC magnetic latching contactor 3 to close, and the DC power supplies the OLT device 10 and charges the OLT backup battery 11 through the DC magnetic latching contactor 3.

[0114] When the mains power detection module 15 detects that there is no mains power, the main control module 5 controls the DC magnetic latching contactor 3 to open, and the operator's DC load 9 is powered by the base station backup battery 8; the OLT equipment 10 is powered by the OLT backup battery 11 alone.

[0115] The main control module 5 is connected to the metering control module 2, the DC magnetic latching contactor 3, the visual recognition module 7, and the mains power detection module 15, respectively.

[0116] There are multiple metering control modules 2; each metering control module 2 is connected to an operator's DC load 9;

[0117] The metering control module 2 is used to monitor the load status of the DC load 9 of the operator connected to it in real time, and transmit the monitored load status to the main control module 5; the load status includes voltage, current and power.

[0118] The main control module 5 determines whether an abnormal load situation has occurred based on the transmitted load status. If an abnormal load situation occurs, it triggers an alarm or controls the metering control module 2 to perform protective measures.

[0119] The visual recognition module 7 is connected to the camera 12, the infrared sensor 13, and the ultrasonic sensor 14 respectively.

[0120] Camera 12 is used for image acquisition at the base station;

[0121] Infrared sensor 13 is used for real-time acquisition of thermal images of the base station;

[0122] The ultrasonic sensor 14 is used to collect ultrasonic data from the base station;

[0123] When the infrared sensor 13 detects a change in the thermal image due to a person or animal entering the base station, or when the ultrasonic sensor 14 detects ultrasonic waves caused by the operation of equipment outside the base station, the camera 12 is awakened; otherwise, the camera 12 is in sleep mode.

[0124] The visual recognition module 7 is used to determine whether a person or animal has entered the base station based on the data collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14. If it is an animal, an alarm is triggered. If it is a person entering the base station, it is necessary to determine whether the person has access permission. If the person does not have access permission, an alarm is triggered. If the person has access permission, it is further determined whether the person has abnormal behavior. If abnormal behavior is found, an alarm is triggered.

[0125] The visual recognition module 7 sends the content collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14, along with the judgment made by the visual recognition module 7, to the main control module 5.

[0126] The main control module 5 sends its interaction content with the metering control module 2 and the visual recognition module 7, as well as the status of the DC magnetic holding contactor 3 and the detection results of the mains power detection module 15, to the cloud platform via the 4G communication module 6, to the WeChat mini program via the Bluetooth module 16, and to the tower operation and maintenance platform via the FSU 17.

[0127] Example 2

[0128] like Figure 2 As shown, a base station DC integrated management and control device includes a switching power supply 1, a metering control module 2, a DC magnetic latching contactor 3, a Schottky diode 4, a main control module 5, a 4G communication module 6, a visual recognition module 7, a base station backup battery 8, an OLT backup battery 11, a camera 12, an infrared sensor 13, an ultrasonic sensor 14, a mains power detection module 15, a Bluetooth module 16, and an FSU 17.

[0129] Three-phase AC power is connected to switching power supply 1 to output 48V DC power;

[0130] Metering control module 2, OLT device 10, and OLT backup battery 11 are cascaded in sequence;

[0131] The 0V terminal of the switching power supply 1 is connected to the positive terminal of the metering control module 2, one end of the DC magnetic latching contactor 3, and the positive terminal of the Schottky diode 4, respectively.

[0132] The other end of the DC magnetic latching contactor 3 is connected to the negative terminal of the Schottky diode 4, the positive terminal of the OLT device 10, and the positive terminal of the OLT backup battery 11, respectively.

[0133] The -48V terminal of the switching power supply 1 is connected to the negative terminal of the metering control module 2, the negative terminal of the OLT device 10, and the negative terminal of the OLT backup battery 11, respectively.

[0134] The switching power supply 1 is also connected to the base station backup battery 8;

[0135] The mains power detection module 15 is connected to the switching power supply 1 and is used to check whether the input received by the switching power supply 1 is mains power;

[0136] When the mains power detection module 15 detects that the mains power is available, the main control module 5 controls the DC magnetic latching contactor 3 to close, and the DC power supplies the OLT device 10 and charges the OLT backup battery 11 through the DC magnetic latching contactor 3.

[0137] When the mains power detection module 15 detects that there is no mains power, the main control module 5 controls the DC magnetic latching contactor 3 to open, and the operator's DC load 9 is powered by the base station backup battery 8; the OLT equipment 10 is powered by the OLT backup battery 11 alone.

[0138] The main control module 5 is connected to the metering control module 2, the DC magnetic latching contactor 3, the visual recognition module 7, and the mains power detection module 15, respectively.

[0139] There are multiple metering control modules 2; each metering control module 2 is connected to an operator's DC load 9;

[0140] The metering control module 2 is used to monitor the load status of the DC load 9 of the operator connected to it in real time, and transmit the monitored load status to the main control module 5; the load status includes voltage, current and power.

[0141] The main control module 5 determines whether an abnormal load situation has occurred based on the transmitted load status. If an abnormal load situation occurs, it triggers an alarm or controls the metering control module 2 to perform protective measures.

[0142] The visual recognition module 7 is connected to the camera 12, the infrared sensor 13, and the ultrasonic sensor 14 respectively.

[0143] Camera 12 is used for image acquisition at the base station;

[0144] Infrared sensor 13 is used for real-time acquisition of thermal images of the base station;

[0145] The ultrasonic sensor 14 is used to collect ultrasonic data from the base station;

[0146] When the infrared sensor 13 detects a change in the thermal image due to a person or animal entering the base station, or when the ultrasonic sensor 14 detects ultrasonic waves caused by the operation of equipment outside the base station, the camera 12 is awakened; otherwise, the camera 12 is in sleep mode.

[0147] The visual recognition module 7 is used to determine whether a person or animal has entered the base station based on the data collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14. If it is an animal, an alarm is triggered. If it is a person entering the base station, it is necessary to determine whether the person has access permission. If the person does not have access permission, an alarm is triggered. If the person has access permission, it is further determined whether the person has abnormal behavior. If abnormal behavior is found, an alarm is triggered.

[0148] The visual recognition module 7 sends the content collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14, along with the judgment made by the visual recognition module 7, to the main control module 5.

[0149] The main control module 5 sends its interaction content with the metering control module 2 and the visual recognition module 7, as well as the status of the DC magnetic holding contactor 3 and the detection results of the mains power detection module 15, to the cloud platform via the 4G communication module 6, to the WeChat mini program via the Bluetooth module 16, and to the tower operation and maintenance platform via the FSU 17.

[0150] It also includes a lightning protection and surge absorption module 18, and the lightning protection and surge absorption module 18, metering and control module 2, OLT equipment 10, and OLT backup battery 11 are cascaded in sequence.

[0151] The switching power supply 1 outputs 48V DC power, which is then supplied to the operator's DC load 9 after the high-frequency pulses from the previous stage are suppressed and absorbed by the lightning protection and surge absorption module 18 and then passed through the DC metering and control module 2.

[0152] Example 3

[0153] like Figure 3 As shown, a base station DC integrated management and control device includes a switching power supply 1, a metering control module 2, a DC magnetic latching contactor 3, a Schottky diode 4, a main control module 5, a 4G communication module 6, a visual recognition module 7, a base station backup battery 8, an OLT backup battery 11, a camera 12, an infrared sensor 13, an ultrasonic sensor 14, a mains power detection module 15, a Bluetooth module 16, and an FSU 17.

[0154] Three-phase AC power is connected to switching power supply 1 to output 48V DC power;

[0155] Metering control module 2, OLT device 10, and OLT backup battery 11 are cascaded in sequence;

[0156] The 0V terminal of the switching power supply 1 is connected to the positive terminal of the metering control module 2, one end of the DC magnetic latching contactor 3, and the positive terminal of the Schottky diode 4, respectively.

[0157] The other end of the DC magnetic latching contactor 3 is connected to the negative terminal of the Schottky diode 4, the positive terminal of the OLT device 10, and the positive terminal of the OLT backup battery 11, respectively.

[0158] The -48V terminal of the switching power supply 1 is connected to the negative terminal of the metering control module 2, the negative terminal of the OLT device 10, and the negative terminal of the OLT backup battery 11, respectively.

[0159] The switching power supply 1 is also connected to the base station backup battery 8;

[0160] The mains power detection module 15 is connected to the switching power supply 1 and is used to check whether the input received by the switching power supply 1 is mains power;

[0161] When the mains power detection module 15 detects that the mains power is available, the main control module 5 controls the DC magnetic latching contactor 3 to close, and the DC power supplies the OLT device 10 and charges the OLT backup battery 11 through the DC magnetic latching contactor 3.

[0162] When the mains power detection module 15 detects that there is no mains power, the main control module 5 controls the DC magnetic latching contactor 3 to open, and the operator's DC load 9 is powered by the base station backup battery 8; the OLT equipment 10 is powered by the OLT backup battery 11 alone.

[0163] The main control module 5 is connected to the metering control module 2, the DC magnetic latching contactor 3, the visual recognition module 7, and the mains power detection module 15, respectively.

[0164] There are multiple metering control modules 2; each metering control module 2 is connected to an operator's DC load 9;

[0165] The metering control module 2 is used to monitor the load status of the DC load 9 of the operator connected to it in real time, and transmit the monitored load status to the main control module 5; the load status includes voltage, current and power.

[0166] The main control module 5 determines whether an abnormal load situation has occurred based on the transmitted load status. If an abnormal load situation occurs, it triggers an alarm or controls the metering control module 2 to perform protective measures.

[0167] The visual recognition module 7 is connected to the camera 12, the infrared sensor 13, and the ultrasonic sensor 14 respectively.

[0168] Camera 12 is used for image acquisition at the base station;

[0169] Infrared sensor 13 is used for real-time acquisition of thermal images of the base station;

[0170] The ultrasonic sensor 14 is used to collect ultrasonic data from the base station;

[0171] When the infrared sensor 13 detects a change in the thermal image due to a person or animal entering the base station, or when the ultrasonic sensor 14 detects ultrasonic waves caused by the operation of equipment outside the base station, the camera 12 is awakened; otherwise, the camera 12 is in sleep mode.

[0172] The visual recognition module 7 is used to determine whether a person or animal has entered the base station based on the data collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14. If it is an animal, an alarm is triggered. If it is a person entering the base station, it is necessary to determine whether the person has access permission. If the person does not have access permission, an alarm is triggered. If the person has access permission, it is further determined whether the person has abnormal behavior. If abnormal behavior is found, an alarm is triggered.

[0173] The visual recognition module 7 sends the content collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14, along with the judgment made by the visual recognition module 7, to the main control module 5.

[0174] The main control module 5 sends its interaction content with the metering control module 2 and the visual recognition module 7, as well as the status of the DC magnetic holding contactor 3 and the detection results of the mains power detection module 15, to the cloud platform via the 4G communication module 6, to the WeChat mini program via the Bluetooth module 16, and to the tower operation and maintenance platform via the FSU 17.

[0175] It also includes a lightning protection and surge absorption module 18, and the lightning protection and surge absorption module 18, metering and control module 2, OLT equipment 10, and OLT backup battery 11 are cascaded in sequence.

[0176] The switching power supply 1 outputs 48V DC power, which is then supplied to the operator's DC load 9 after the high-frequency pulses from the previous stage are suppressed and absorbed by the lightning protection and surge absorption module 18 and then passed through the DC metering and control module 2.

[0177] The switching power supply 1 includes an AC / DC module 19 and a secondary low-voltage disconnect contactor 20;

[0178] The three-phase AC power is connected to the AC / DC module 19 of the switching power supply 1 to output 48V DC power. After voltage detection by the secondary low-voltage disconnect contactor 20 of the switching power supply 1, the output can be controlled.

[0179] Example 4

[0180] like Figure 3 As shown, a base station DC integrated management and control device includes a switching power supply 1, a metering control module 2, a DC magnetic latching contactor 3, a Schottky diode 4, a main control module 5, a 4G communication module 6, a visual recognition module 7, a base station backup battery 8, an OLT backup battery 11, a camera 12, an infrared sensor 13, an ultrasonic sensor 14, a mains power detection module 15, a Bluetooth module 16, and an FSU 17.

[0181] Three-phase AC power is connected to switching power supply 1 to output 48V DC power;

[0182] Metering control module 2, OLT device 10, and OLT backup battery 11 are cascaded in sequence;

[0183] The 0V terminal of the switching power supply 1 is connected to the positive terminal of the metering control module 2, one end of the DC magnetic latching contactor 3, and the positive terminal of the Schottky diode 4, respectively.

[0184] The other end of the DC magnetic latching contactor 3 is connected to the negative terminal of the Schottky diode 4, the positive terminal of the OLT device 10, and the positive terminal of the OLT backup battery 11, respectively.

[0185] The -48V terminal of the switching power supply 1 is connected to the negative terminal of the metering control module 2, the negative terminal of the OLT device 10, and the negative terminal of the OLT backup battery 11, respectively.

[0186] The switching power supply 1 is also connected to the base station backup battery 8;

[0187] The mains power detection module 15 is connected to the switching power supply 1 and is used to check whether the input received by the switching power supply 1 is mains power;

[0188] When the mains power detection module 15 detects that the mains power is available, the main control module 5 controls the DC magnetic latching contactor 3 to close, and the DC power supplies the OLT device 10 and charges the OLT backup battery 11 through the DC magnetic latching contactor 3.

[0189] When the mains power detection module 15 detects that there is no mains power, the main control module 5 controls the DC magnetic latching contactor 3 to open, and the operator's DC load 9 is powered by the base station backup battery 8; the OLT equipment 10 is powered by the OLT backup battery 11 alone.

[0190] The main control module 5 is connected to the metering control module 2, the DC magnetic latching contactor 3, the visual recognition module 7, and the mains power detection module 15, respectively.

[0191] There are multiple metering control modules 2; each metering control module 2 is connected to an operator's DC load 9;

[0192] The metering control module 2 is used to monitor the load status of the DC load 9 of the operator connected to it in real time, and transmit the monitored load status to the main control module 5; the load status includes voltage, current and power.

[0193] The main control module 5 determines whether an abnormal load situation has occurred based on the transmitted load status. If an abnormal load situation occurs, it triggers an alarm or controls the metering control module 2 to perform protective measures.

[0194] The visual recognition module 7 is connected to the camera 12, the infrared sensor 13, and the ultrasonic sensor 14 respectively.

[0195] Camera 12 is used for image acquisition at the base station;

[0196] Infrared sensor 13 is used for real-time acquisition of thermal images of the base station;

[0197] The ultrasonic sensor 14 is used to collect ultrasonic data from the base station;

[0198] When the infrared sensor 13 detects a change in the thermal image due to a person or animal entering the base station, or when the ultrasonic sensor 14 detects ultrasonic waves caused by the operation of equipment outside the base station, the camera 12 is awakened; otherwise, the camera 12 is in sleep mode.

[0199] The visual recognition module 7 is used to determine whether a person or animal has entered the base station based on the data collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14. If it is an animal, an alarm is triggered. If it is a person entering the base station, it is necessary to determine whether the person has access permission. If the person does not have access permission, an alarm is triggered. If the person has access permission, it is further determined whether the person has abnormal behavior. If abnormal behavior is found, an alarm is triggered.

[0200] The visual recognition module 7 sends the content collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14, along with the judgment made by the visual recognition module 7, to the main control module 5.

[0201] The main control module 5 sends its interaction content with the metering control module 2 and the visual recognition module 7, as well as the status of the DC magnetic holding contactor 3 and the detection results of the mains power detection module 15, to the cloud platform via the 4G communication module 6, to the WeChat mini program via the Bluetooth module 16, and to the tower operation and maintenance platform via the FSU 17.

[0202] It also includes a lightning protection and surge absorption module 18, and the lightning protection and surge absorption module 18, metering and control module 2, OLT equipment 10, and OLT backup battery 11 are cascaded in sequence.

[0203] The switching power supply 1 outputs 48V DC power, which is then supplied to the operator's DC load 9 after the high-frequency pulses from the previous stage are suppressed and absorbed by the lightning protection and surge absorption module 18 and then passed through the DC metering and control module 2.

[0204] The switching power supply 1 includes an AC / DC module 19 and a secondary low-voltage disconnect contactor 20;

[0205] The three-phase AC power is connected to the AC / DC module 19 of the switching power supply 1 to output 48V DC power. After voltage detection by the secondary low-voltage disconnect contactor 20 of the switching power supply 1, the output can be controlled.

[0206] When the OLT backup battery 11 fails or is not connected, the OLT device 10 is powered by the base station backup battery 8.

[0207] In the main control module 5, the abnormal load conditions include overcurrent, undervoltage, and overload; when overcurrent or undervoltage occurs, an alarm is triggered; when overload occurs, the metering control module 2 is controlled to perform protection measures; the protection measures are to disconnect the DC output and automatically restore the DC output after the abnormality is resolved.

[0208] The main control module 5 is also used to calculate the power consumption of each operator's DC load 9 based on the transmitted load status. At the same time, the main control module 5 distinguishes the power consumption according to time, dividing it into the power consumption of each operator today, yesterday, this month, last month, and total power consumption, as well as the total power consumption of this device and yesterday's power consumption. The main control module 5 multiplies the power consumption of each operator today, yesterday, this month, last month, and total power consumption by the preset electricity price to obtain the electricity cost data of each operator. After storing the above data locally, the main control module 5 sends it to the FSU 17 and 4G communication module 6 through the RS485 interface. After receiving the data, the FSU 17 and 4G communication module 6 send it to the tower operation and maintenance platform and the cloud platform, respectively. The tower operation and maintenance platform and the cloud platform display the data and export reports based on the received data, which serve as the original basis for electricity cost settlement.

[0209] The main control module 5 processes the interaction content between the pre-defined standard wired and wireless protocol texts and the metering control module 2, the visual recognition module 7, the status of the DC magnetic holding contactor 3, and the detection results of the mains power detection module 15, and then sends them to the 4G communication module 6, the Bluetooth module 16, and the FSU 17. After parsing, the FSU 17 uploads the data to the tower operation and maintenance platform through its built-in wireless module.

[0210] The 4G communication module 6 sends data to the cloud platform;

[0211] After the Bluetooth module 16 establishes a connection with the WeChat mini-program on the mobile phone, it sends data to the WeChat mini-program to achieve local data monitoring.

[0212] The other end of the DC magnetic latching contactor 3 is also connected to the positive terminal of the camera 12, and the -48V terminal of the switching power supply 1 is also connected to the negative terminal of the camera 12.

[0213] The visual recognition module 7 has a built-in face recognition and behavior analysis model;

[0214] The face recognition and behavior analysis model includes a hierarchical feature extraction module, an adaptive fusion module, a key point feature extraction module, an adaptive convolution module, a multi-level sparse coding module, and a behavior recognition and temporal processing module;

[0215] The hierarchical feature extraction module uses multi-scale pyramid decomposition and then lightweight convolution to convolve the image at each scale to obtain feature maps at different levels.

[0216] The adaptive fusion module is used to fuse feature maps at different levels to obtain a fused feature map.

[0217] The key point feature extraction module is used to detect key points in all feature maps to achieve face recognition;

[0218] The adaptive convolution module is used to adaptively adjust the convolution operation according to the complexity of the image, retaining only the most important convolution kernels;

[0219] The multi-level sparse coding module is used to obtain the sparse coding vectors of all feature maps and merge them to obtain the sparsed feature vectors at each time step.

[0220] The behavior recognition and temporal processing module is used to perform behavior recognition based on the sparse feature vectors at each time step.

[0221] Application Examples

[0222] like Figure 3 As shown, a base station DC integrated management and control device includes a switching power supply 1, a metering control module 2, a DC magnetic latching contactor 3, a Schottky diode 4, a main control module 5, a 4G communication module 6, a visual recognition module 7, a base station backup battery 8, an OLT backup battery 11, a camera 12, an infrared sensor 13, an ultrasonic sensor 14, a mains power detection module 15, a Bluetooth module 16, and an FSU 17.

[0223] Three-phase AC power is connected to switching power supply 1 to output 48V DC power;

[0224] Metering control module 2, OLT device 10, and OLT backup battery 11 are cascaded in sequence;

[0225] The 0V terminal of the switching power supply 1 is connected to the positive terminal of the metering control module 2, one end of the DC magnetic latching contactor 3, and the positive terminal of the Schottky diode 4, respectively.

[0226] The other end of the DC magnetic latching contactor 3 is connected to the negative terminal of the Schottky diode 4, the positive terminal of the OLT device 10, and the positive terminal of the OLT backup battery 11, respectively.

[0227] The -48V terminal of the switching power supply 1 is connected to the negative terminal of the metering control module 2, the negative terminal of the OLT device 10, and the negative terminal of the OLT backup battery 11, respectively.

[0228] The switching power supply 1 is also connected to the base station backup battery 8;

[0229] The mains power detection module 15 is connected to the switching power supply 1 and is used to check whether the input received by the switching power supply 1 is mains power;

[0230] When the mains power detection module 15 detects that the mains power is available, the main control module 5 controls the DC magnetic latching contactor 3 to close, and the DC power supplies the OLT device 10 and charges the OLT backup battery 11 through the DC magnetic latching contactor 3.

[0231] When the mains power detection module 15 detects that there is no mains power, the main control module 5 controls the DC magnetic latching contactor 3 to open, and the operator's DC load 9 is powered by the base station backup battery 8; the OLT equipment 10 is powered by the OLT backup battery 11 alone.

[0232] The main control module 5 is connected to the metering control module 2, the DC magnetic latching contactor 3, the visual recognition module 7, and the mains power detection module 15, respectively.

[0233] There are multiple metering control modules 2; each metering control module 2 is connected to an operator's DC load 9;

[0234] The metering control module 2 is used to monitor the load status of the DC load 9 of the operator connected to it in real time, and transmit the monitored load status to the main control module 5; the load status includes voltage, current and power.

[0235] The main control module 5 determines whether an abnormal load situation has occurred based on the transmitted load status. If an abnormal load situation occurs, it triggers an alarm or controls the metering control module 2 to perform protective measures.

[0236] The visual recognition module 7 is connected to the camera 12, the infrared sensor 13, and the ultrasonic sensor 14 respectively.

[0237] Camera 12 is used for image acquisition at the base station;

[0238] Infrared sensor 13 is used for real-time acquisition of thermal images of the base station;

[0239] The ultrasonic sensor 14 is used to collect ultrasonic data from the base station;

[0240] When the infrared sensor 13 detects a change in the thermal image due to a person or animal entering the base station, or when the ultrasonic sensor 14 detects ultrasonic waves caused by the operation of equipment outside the base station, the camera 12 is awakened; otherwise, the camera 12 is in sleep mode.

[0241] The visual recognition module 7 is used to determine whether a person or animal has entered the base station based on the data collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14. If it is an animal, an alarm is triggered. If it is a person entering the base station, it is necessary to determine whether the person has access permission. If the person does not have access permission, an alarm is triggered. If the person has access permission, it is further determined whether the person has abnormal behavior. If abnormal behavior is found, an alarm is triggered.

[0242] The visual recognition module 7 sends the content collected by the camera 12, infrared sensor 13, and ultrasonic sensor 14, along with the judgment made by the visual recognition module 7, to the main control module 5.

[0243] The main control module 5 sends its interaction content with the metering control module 2 and the visual recognition module 7, as well as the status of the DC magnetic holding contactor 3 and the detection results of the mains power detection module 15, to the cloud platform via the 4G communication module 6, to the WeChat mini program via the Bluetooth module 16, and to the tower operation and maintenance platform via the FSU 17.

[0244] It also includes a lightning protection and surge absorption module 18, and the lightning protection and surge absorption module 18, metering and control module 2, OLT equipment 10, and OLT backup battery 11 are cascaded in sequence.

[0245] The switching power supply 1 outputs 48V DC power, which is then supplied to the operator's DC load 9 after the high-frequency pulses from the previous stage are suppressed and absorbed by the lightning protection and surge absorption module 18 and then passed through the DC metering and control module 2.

[0246] The switching power supply 1 includes an AC / DC module 19 and a secondary low-voltage disconnect contactor 20;

[0247] The three-phase AC power is connected to the AC / DC module 19 of the switching power supply 1 to output 48V DC power. After voltage detection by the secondary low-voltage disconnect contactor 20 of the switching power supply 1, the output can be controlled.

[0248] When the OLT backup battery 11 fails or is not connected, the OLT device 10 is powered by the base station backup battery 8.

[0249] In the main control module 5, the abnormal load conditions include overcurrent, undervoltage, and overload; when overcurrent or undervoltage occurs, an alarm is triggered; when overload occurs, the metering control module 2 is controlled to perform protection measures; the protection measures are to disconnect the DC output and automatically restore the DC output after the abnormality is resolved.

[0250] The main control module 5 is also used to calculate the power consumption of each operator's DC load 9 based on the transmitted load status. At the same time, the main control module 5 distinguishes the power consumption according to time, dividing it into the power consumption of each operator today, yesterday, this month, last month, and total power consumption, as well as the total power consumption of this device and yesterday's power consumption. The main control module 5 multiplies the power consumption of each operator today, yesterday, this month, last month, and total power consumption by the preset electricity price to obtain the electricity cost data of each operator. After storing the above data locally, the main control module 5 sends it to the FSU 17 and 4G communication module 6 through the RS485 interface. After receiving the data, the FSU 17 and 4G communication module 6 send it to the tower operation and maintenance platform and the cloud platform, respectively. The tower operation and maintenance platform and the cloud platform display the data and export reports based on the received data, which serve as the original basis for electricity cost settlement.

[0251] The main control module 5 processes the interaction content between the pre-defined standard wired and wireless protocol texts and the metering control module 2, the visual recognition module 7, the status of the DC magnetic holding contactor 3, and the detection results of the mains power detection module 15, and then sends them to the 4G communication module 6, the Bluetooth module 16, and the FSU 17. After parsing, the FSU 17 uploads the data to the tower operation and maintenance platform through its built-in wireless module.

[0252] The 4G communication module 6 sends data to the cloud platform;

[0253] After the Bluetooth module 16 establishes a connection with the WeChat mini-program on the mobile phone, it sends data to the WeChat mini-program to achieve local data monitoring.

[0254] The other end of the DC magnetic latching contactor 3 is also connected to the positive terminal of the camera 12, and the -48V terminal of the switching power supply 1 is also connected to the negative terminal of the camera 12.

[0255] The visual recognition module 7 has a built-in face recognition and behavior analysis model;

[0256] The face recognition and behavior analysis model includes a hierarchical feature extraction module, an adaptive fusion module, a key point feature extraction module, an adaptive convolution module, a multi-level sparse coding module, and a behavior recognition and temporal processing module;

[0257] The hierarchical feature extraction module uses multi-scale pyramid decomposition and then lightweight convolution to convolve the image at each scale to obtain feature maps at different levels.

[0258] The adaptive fusion module is used to fuse feature maps at different levels to obtain a fused feature map.

[0259] The key point feature extraction module is used to detect key points in all feature maps to achieve face recognition;

[0260] The adaptive convolution module is used to adaptively adjust the convolution operation according to the complexity of the image, retaining only the most important convolution kernels;

[0261] The multi-level sparse coding module is used to obtain the sparse coding vectors of all feature maps and merge them to obtain the sparsed feature vectors at each time step.

[0262] The behavior recognition and temporal processing module is used to perform behavior recognition based on the sparse feature vectors at each time step.

[0263] This example has 10 metering control modules 2; each metering control module 2 is connected to one operator DC load 9, that is, there are 10 operator DC loads 9; therefore, this example can support 10 30A branch controllable DC outputs.

[0264] The OLT backup battery 11 uses a set of 48V 100AH ​​lithium iron phosphate batteries.

[0265] Camera 12 can be a camera with a resolution of 200W pixels or higher.

[0266] The visual recognition module 7 preferably uses the K210 chip.

[0267] This invention's face recognition and behavior analysis model comprises six core modules: a hierarchical feature extraction module, an adaptive fusion module, a key point feature extraction module, an adaptive convolution module (ACM), a multi-level sparse coding module, and a behavior recognition and temporal processing module, enabling efficient analysis and recognition of static face features and dynamic behavior sequences. The connection relationships and data flows of each module are as follows:

[0268] Input: Raw image (single frame or multiple consecutive frames).

[0269] Hierarchical feature extraction module → Adaptive fusion module → Key point feature extraction module → Adaptive convolution module → Multi-level sparse coding module → Behavior recognition and temporal processing module → Output.

[0270] Specifically as follows:

[0271] 1. Hierarchical Feature Extraction Module

[0272] (1) Multi-scale feature generation:

[0273] The input image is subjected to multi-scale pyramid decomposition to obtain image levels with different resolutions. The resolution of each level is scaled down proportionally to the size of the original image (e.g., 1 / 2, 1 / 4, 1 / 8).

[0274] Lightweight convolutions (depth-separable convolutions) are used to convolve the image at each scale to extract basic features;

[0275]

[0276] (x,y): The first The feature map of the layer at location The value at that point represents the feature map obtained after the convolution operation.

[0277] (x,y): The first The input image of the layer is located at Pixel value at;

[0278] : Convolution kernel at position The weight value at the location;

[0279] The size of the convolution kernel (width and height) is usually an odd number, such as 3×3 or 5×5.

[0280] : The index inside the convolution kernel, used to traverse each position of the convolution kernel; and These represent the indices of the convolution kernel in the horizontal and vertical directions, respectively.

[0281] : Position coordinates in the feature map or input image, representing the pixel position in the currently calculated feature map or image.

[0282] : Hierarchical index, representing different levels after multi-scale pyramid decomposition, for example =0 indicates the original image. =1 indicates an image reduced by half, and so on.

[0283] Assuming the input image is 256×256, after multi-scale pyramid decomposition, three levels are obtained:

[0284] =0: Original image 256×256.

[0285] =1: Image reduced to half size, 128×128.

[0286] =2: Reduce the image size by one-quarter to 64×64.

[0287] For each level, a 3×3 convolution kernel is used to perform a convolution operation to extract feature maps. (x,y).

[0288] 2. Adaptive Fusion Module

[0289] An adaptive weighting mechanism is used to fuse feature maps at different levels. The adaptive weights are automatically adjusted based on the image content and complexity to prioritize the fusion of the most important information.

[0290] Adaptive weight calculation formula:

[0291]

[0292] : No. The fusion weights of the layer feature maps represent the weights of the first layer. The importance of layer feature maps in the fusion process is such that the greater the weight, the greater the contribution of that layer feature map to the final fusion result.

[0293] Feature map The complexity and importance evaluation function is used to evaluate the complexity and importance of the first... The complexity of a layer feature map is usually based on the gradient magnitude, information entropy, or other statistical properties of the feature map.

[0294] : No. The feature map of the layer represents the result after the convolution operation. Layer feature map.

[0295] : The hierarchical index of the feature map, representing the set of all feature layers involved in the fusion; The value range is from 1 to L.

[0296] : No. The feature map of the layer represents the result after the convolution operation. Layer feature maps. The feature maps of each layer may have different resolutions and semantic information.

[0297] L: The total number of feature map layers, representing the total number of feature maps involved in the fusion.

[0298] Exponential functions represent exponential operations with the natural constant e as the base, and are used to score complexity. Convert to positive numbers and calculate the fusion weights through normalization.

[0299] 3. Key Point Feature Extraction Module

[0300] Keypoint detection is performed on the feature map, specifically for key features such as eyes, nose, mouth, and eyebrows. A lightweight model (such as MobileNetV2) is then used to extract local features from these keypoints, resulting in a local feature vector for face recognition.

[0301] Key point feature extraction formula:

[0302]

[0303] Key points The local feature vector at the key point represents the local feature vector at the key point. The local features extracted at the point are used to describe the feature information of the key point.

[0304] : No. Layer feature map at location The value at that location represents the pixel value of the area surrounding the keypoint, used to extract local features.

[0305] The convolutional kernel used to extract local features is located at... The weights at each location represent the weights of the convolution kernels used to extract local features.

[0306] The radius of the convolution kernel represents its size. The width and height of the convolution kernel are... .

[0307] : The coordinates of the key point in the feature map, indicating the location of the key point for extracting local features.

[0308] The process of key point feature extraction:

[0309] (1) Input: the first Feature map of layer and key locations .

[0310] (2) Local feature extraction:

[0311] Key points Extract the surrounding area as the center ( )×( Characteristics of the region;

[0312] Using convolution kernels By performing a weighted summation on the local regions, the local feature vectors of the key points are obtained. .

[0313] (3) Output: Key Points Local feature vectors at [location] .

[0314] 4. Adaptive Convolutional Module (ACM)

[0315] (1) Dynamic convolution selection:

[0316] The adaptive convolution module adjusts the convolution operation adaptively based on the complexity of the input image. If complex features are detected in the input image, the adaptive convolution module will enable multiple convolution modules and increase the number of convolution kernels; if simple features are detected, the convolution structure will be simplified and only the basic convolution module will be enabled.

[0317] Dynamic convolution selection formula:

[0318]

[0319] The adaptive convolution module selects the convolution operation; it scores based on the complexity of the input image. Choose deep convolution ( or shallow convolution ( ).

[0320] : Complexity score of the input image, used to measure the complexity of the input image, usually calculated based on global gradient and edge density.

[0321] Complexity threshold: Used to determine whether the complexity of the input image exceeds the threshold, thereby deciding whether to choose deep convolution or shallow convolution.

[0322] Deep convolution operations use multiple layers of convolution kernels and more channels, making them suitable for complex images.

[0323] Shallow convolution operations use a single convolution kernel and fewer channels, making them suitable for simple images.

[0324] Calculation:

[0325] 1) Global gradient: Calculate the global gradient magnitude of the input image. The larger the gradient magnitude, the more detail the image contains. Formula:

[0326]

[0327] in, It is an image The number of pixels at the middle edge.

[0328] 2) Edge density:

[0329] Calculate the edge density (EdgeDensity(I)) of the image. A higher edge density indicates that the image contains more edge information. Formula:

[0330] EdgeDensity(I) =

[0331] Here, EdgePixelCount(I) is the number of edge pixels in image I.

[0332] 3) Complexity score:

[0333] The complexity score is calculated by combining global gradient and edge density. :

[0334]

[0335] in, and It is a weighting parameter used to balance the effects of gradient and edge density.

[0336] (2) Sparse weight update:

[0337] During convolution, the adaptive convolution module sparsifies the weight matrix, retaining only the most important convolution kernels while zeroing out the weights of other kernels, thereby reducing computation and memory usage.

[0338] Sparsification formula:

[0339]

[0340] : The sparsified weight matrix, representing the convolution kernel weight matrix after sparsification.

[0341] : Original weight matrix, representing the original weights of the convolution kernel.

[0342] : Sparsity threshold, used to determine whether a weight is important; greater than The weight is retained, less than or equal to The weights are zeroed out.

[0343] : Index of the weight matrix, indicating its position in the weight matrix.

[0344] Among them, for the sparse threshold Adjustment: Sparse threshold It can be dynamically adjusted based on the complexity of the input features.

[0345] For example:

[0346]

[0347] in:

[0348] The mean of the weight matrix W;

[0349] : Standard deviation of the weight matrix W;

[0350] 5. Multilevel Sparse Coding Module (MSC)

[0351] (1) Multi-level application of sparse dictionaries

[0352]

[0353] : No. A sparse dictionary of layer features, representing the features used to encode the first layer. The sparse dictionary matrix of the layer feature map.

[0354] : No. The first layer of the sparse dictionary Each atom represents a basis vector in the sparse dictionary, used to represent key information in the feature map.

[0355] n: The number of atoms in the sparse dictionary, representing the dimension of the sparse dictionary.

[0356] Sparse dictionary It can be obtained through the following methods:

[0357] Pre-training: Use a sparse coding algorithm (such as K-SVD) to pre-train the training data to obtain a sparse dictionary.

[0358] (2) Sparse decoding and merging

[0359]

[0360] : No. The feature map of the layer; representing the feature map of the first layer after the convolution operation. Feature map.

[0361] : No. The sparse coding vector of the layer feature map represents the feature map. In sparse dictionaries Sparse representation below.

[0362] : Sparsity adjustment parameter, controls the sparsity of sparse coding; the larger the value of λ, the stronger the sparsity of the encoded vector. The sparser.

[0363] The square of the L2 norm represents the reconstruction error and is used to measure the feature map. With sparse representation The differences between them.

[0364] L1 norm, representing a sparse coding vector. Sparsity.

[0365] Solving sparse coding: Sparse coding vector The solution can be obtained using the Iterative Soft Thresholding Algorithm (ISTA).

[0366] (3) Merge the coding results of all levels

[0367] Merging sparse coding vectors from all levels This generates the final low-dimensional, sparse feature vector. :

[0368]

[0369] in, It is the first The weights of layer features can be dynamically adjusted based on the importance of the features.

[0370] 6. Behavior recognition and timing processing module

[0371] (1) Time series modeling:

[0372] Multi-frame feature encoding results , ···, Combined into a time series; Let be the sparsed feature vector of the frame number at time T.

[0373] Modeling time series information using a lightweight LSTM:

[0374]

[0375] in, It is the hidden state of time step t.

[0376] Behavior recognition: based on the hidden state of LSTM It classifies and identifies continuous actions.

[0377] The face recognition and behavior analysis model used in this device was compared with commonly used algorithms in the industry. During the comparison, a low-power embedded platform (ESP32 and ARM Cortex-M7 MCU) was used for testing; the results are shown in Table 1.

[0378] Table 1

[0379]

[0380] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A base station DC integrated control device, characterized in that: The lightning protection and surge absorption module, metering and control module, OLT equipment, and OLT backup battery are cascaded in sequence; The 0V terminal of the switching power supply is connected to the positive terminals of multiple metering and control modules, one end of the DC magnetic latching contactor, and the positive terminal of the Schottky diode, respectively. The other end of the DC magnetic latching contactor is connected to the negative terminal of the Schottky diode, the positive terminal of the OLT equipment, and the positive terminal of the OLT backup battery, respectively. The -48V terminal of the switching power supply is connected to the negative terminal of the metering control module, the negative terminal of the OLT equipment, and the negative terminal of the OLT backup battery, respectively. The switching power supply is also connected to the base station's backup battery; The mains power detection module is connected to the switching power supply. When the mains power detection module detects no mains power, the operator's DC load is powered by the base station's backup battery; the OLT equipment is powered by the OLT backup battery alone. Each metering control module is connected to a DC load of an operator to monitor the load status of the connected DC load in real time; the main control module determines whether an abnormality has occurred based on the load status, and if so, triggers an alarm or controls the metering control module to perform protective measures. The camera is activated when the infrared sensor detects changes in the thermal image caused by a person or animal entering the base station, or when the ultrasonic sensor detects ultrasonic waves caused by the operation of equipment outside the base station. The visual recognition module determines whether a person or animal has entered the base station based on the data collected by the camera, infrared sensor, and ultrasonic sensor. If it is an animal, an alarm is triggered. If a person enters the base station, it will determine whether they have access privileges. If not, an alarm will be triggered. If so, determine whether there is any abnormal behavior; if there is abnormal behavior, trigger an alarm; then send the collected results and the determined situation to the main control module. The visual recognition module incorporates a face recognition and behavior analysis model, which includes a hierarchical feature extraction module, an adaptive fusion module, a key point feature extraction module, an adaptive convolution module, a multi-level sparse coding module, and a behavior recognition and temporal processing module. The hierarchical feature extraction module uses multi-scale pyramid decomposition and then uses lightweight convolution to convolve the image at each scale to obtain feature maps at different levels. The adaptive fusion module is used to fuse feature maps at different levels to obtain a fused feature map. The key point feature extraction module is used to detect key points in all feature maps to achieve face recognition; The adaptive convolution module is used to adaptively adjust the convolution operation according to the complexity of the image, retaining only the most important convolution kernels; The multi-level sparse coding module is used to obtain the sparse coding vectors of all feature maps and merge them to obtain the sparsed feature vectors at each time step. The behavior recognition and temporal processing module is used to perform behavior recognition based on the sparse feature vectors at each time step.

2. The base station DC integrated management and control device according to claim 1, characterized in that, The switching power supply includes an AC / DC module and a two-stage low-voltage disconnect contactor; The three-phase AC power is connected to the AC / DC module of the switching power supply, which outputs 48V DC power. After voltage detection by the secondary low-voltage disconnect contactor of the switching power supply, a controllable output is achieved.

3. The base station DC integrated management and control device according to claim 1, characterized in that, When the OLT backup battery fails or is not connected, the OLT equipment is powered by the base station backup battery. The main control module sends its interactions with the metering control module and the visual recognition module, as well as the status of the DC magnetic holding contactor and the detection results of the mains power detection module, to the cloud platform via the 4G communication module, to the WeChat mini-program via the Bluetooth module, and to the tower operation and maintenance platform via the FSU.

4. The base station DC integrated management and control device according to claim 1, characterized in that, The abnormal load conditions include overcurrent, undervoltage, and overload; when overcurrent or undervoltage occurs, an alarm is triggered. When an overload occurs, the metering control module will execute protective measures; the protective measures are to disconnect the DC output and automatically restore the DC output once the abnormality is resolved.

5. The base station DC integrated management and control device according to claim 1, characterized in that, The main control module is also used to calculate the power consumption of each operator's DC load based on the transmitted load status. Simultaneously, the main control module distinguishes power consumption by time, dividing it into today's power consumption, yesterday's power consumption, this month's power consumption, last month's power consumption, total power consumption, and the device's total power consumption and yesterday's power consumption. The main control module multiplies each operator's today's power consumption, yesterday's power consumption, this month's power consumption, last month's power consumption, and total power consumption by a pre-set electricity price to obtain the electricity cost data for each operator. After storing the above data locally, the main control module sends it to the FSU and 4G communication modules via the RS485 interface. Upon receiving the data, the FSU and 4G communication modules send it to the tower maintenance platform and cloud platform, respectively. Both the tower maintenance platform and cloud platform display the received data and export reports, serving as the original basis for electricity cost settlement.

6. The base station DC integrated management and control device according to claim 1, characterized in that, The main control module processes the interaction content between the pre-defined standard wired and wireless protocol texts and the metering control module, the visual recognition module, the status of the DC magnetic latching contactor, and the detection results of the mains power detection module before sending them to the 4G communication module, Bluetooth module, and FSU. After parsing, the FSU uploads the data to the tower operation and maintenance platform through its built-in wireless module. The 4G communication module sends data to the cloud platform; After the Bluetooth module establishes a connection with the WeChat mini-program on the mobile phone, it sends data to the WeChat mini-program to achieve local data monitoring.

7. The base station DC integrated management and control device according to claim 1, characterized in that, The other end of the DC magnetic latching contactor is also connected to the positive terminal of the camera, and the -48V terminal of the switching power supply is also connected to the negative terminal of the camera.

Images