A method for redundant switching control of power supply for unmanned aerial vehicles (UAVs)
By establishing a two-way encrypted control link and a distributed power status database, multi-dimensional power supply parameters are collected in real time, and the health status is quantitatively assessed. The backup power switching is controlled by a slow start timing control, which solves the problem of balancing power supply reliability and utilization in the power redundancy switching scheme for UAVs. This achieves uninterrupted power supply switching and reliable data storage, and reduces operating costs.
Patent Information
- Application Number
- CN202611453286.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-18
- Publication Date
- 2026-11-17
AI Technical Summary
Existing power redundancy switching solutions for drones cannot simultaneously adapt to the power supply needs of different priority tasks, making it difficult to balance power supply reliability and power utilization, and thus failing to meet the cost reduction and efficiency improvement requirements in the process of large-scale drone applications.
By establishing a two-way encrypted control link, building a distributed power status database, collecting multi-dimensional power supply operation parameters in real time, quantitatively assessing the health of the main power unit, and adopting a soft-start timing control for backup power switching, uninterrupted power supply switching is achieved, and full-process data is archived and stored.
It enables uninterrupted power supply switching, reduces false switching rate, improves power utilization, ensures power supply stability, provides reliable data support, and reduces operating costs.
Smart Images

Figure CN123121971A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of unmanned aerial vehicle (UAV) power control technology, and in particular to a method for redundant switching control of UAV power supply. Background Technology
[0002] As the core carrier for low-altitude operations, drones are now widely used in various fields such as high-voltage power transmission line inspection, urban last-mile logistics delivery, emergency mapping, and plant protection operations. The operation scenarios cover high-altitude mountainous areas, dense urban airspace, and complex field environments. The reliability of their power supply system directly determines the completion rate of the operation and flight safety. The core requirement of the industry is to achieve uninterrupted and stable power supply while taking into account power utilization and operation and maintenance efficiency, and controlling the overall operating costs.
[0003] Currently, mainstream drone power redundancy switching solutions fall into two categories. The first is the relay hard switching solution, which works by directly controlling the switching between the main power supply and the backup power supply circuit through the on / off state of a relay. This is achieved by preseting a single voltage threshold. When the main power supply output voltage falls below the threshold, the relay is triggered. This solution has a simple structure and low hardware cost, and is often used in small drones with low payload and low priority operations. However, this solution involves millisecond-level power interruptions during the switching process, and large voltage spikes are generated at the moment of switching on and off. This can easily impact sensitive electronic components such as the flight control system and precision inspection payloads, and in severe cases, it can cause flight control system crashes, payload data loss, or even crashes.
[0004] The second type is the dual-power parallel hot standby switching solution. Its working principle involves always connecting the main power supply and the backup power supply in parallel to the power supply circuit. When the main power supply is at full load, the backup power supply remains idle and ready to operate. When the main power supply's output power is insufficient, the backup power supply automatically supplements the supply. This solution offers fast switching speed and can essentially achieve uninterrupted power supply, making it suitable for high-value, high-priority industrial drones. However, this solution does not consider multi-dimensional assessment of power supply health, relying solely on a single voltage threshold to trigger switching. This makes it prone to false switching due to instantaneous voltage fluctuations. Furthermore, the backup power supply remains idle and ready for operation for extended periods, resulting in low overall power utilization. Additionally, the lack of a unified power supply status data storage and archiving mechanism makes it impossible to provide data support for subsequent power supply fault tracing and maintenance strategy optimization.
[0005] The current two types of solutions cannot simultaneously adapt to the power supply requirements of different priority tasks, making it difficult to balance power supply reliability and power utilization. They cannot meet the cost reduction and efficiency improvement requirements in the process of large-scale application of drones. There is an urgent need for a highly reliable power redundancy switching solution that can adapt to multiple scenarios to solve the above problems. Summary of the Invention
[0006] This invention proposes a power supply redundancy switching control method for unmanned aerial vehicles (UAVs) to solve the problems mentioned in the prior art.
[0007] To achieve the above objectives, the present invention adopts the following technical solution: a method for redundant switching control of power supply for unmanned aerial vehicles (UAVs), comprising the following steps: Relying on the communication components of the power supply system, a two-way encrypted control link is established between each redundant power supply unit, status acquisition terminal and power supply control backend. The identity, rated parameters and historical operating data of each redundant power supply unit are hash-verified and stored in the distributed power supply status database. Based on the reference parameters of the same type of power supply in the database, the basic power supply data calibration is completed. The status acquisition terminal collects multi-dimensional power supply operation parameters of each redundant power supply unit in real time according to a dynamically adjustable preset sampling period. After preliminary preprocessing of the collected data, it transmits the data to the power supply control backend through an encrypted control link and updates the real-time data entries of the distributed power supply status database synchronously. The power supply control backend retrieves real-time operating data and historical fault sample data from the distributed power supply status database, quantitatively assesses the health of the main power supply unit, adjusts the health switching threshold according to the power supply task priority, and sends a redundant power supply switching command to the switching control execution terminal when the health is lower than the switching threshold. After the switching control execution terminal verifies the switching command, it controls the power supply circuit of the backup power unit to be switched on to be turned on according to the preset soft start sequence. After the output voltage stabilizes in the rated range, it controls the power supply circuit of the original main power unit to be disconnected, completing the uninterrupted power supply switching. At the same time, all data of the switching process is uploaded to the distributed power status database for archiving and storage.
[0008] Furthermore, it also includes the step of real-time operation and maintenance of the distributed power status database. The database health status is calculated using the following formula: ; in, A health score is assigned to the distributed power state database. For power supply data integrity weighting coefficient, This refers to the proportion of valid power supply data entries in the database out of the total number of data entries. The power supply data query response efficiency weighting coefficient. The power supply data query response efficiency is scored. For power supply data storage security weighting coefficient, Assess the security of power supply data storage. Correction factor for abnormal power supply data. This represents the percentage of unarchived abnormal power supply fault data entries in the database. The three satisfy the constraint that the sum of their weights is 1.
[0009] Furthermore, it also includes the step of real-time monitoring of the standby status of each redundant power supply unit, periodically issuing status query commands to the standby power supply unit, collecting real-time power supply operation parameters, conducting a pre-assessment of the health of the standby power supply unit based on the collected parameters, generating alarm information when the pre-assessed health is lower than the standby threshold, synchronously updating the standby power supply status identifier in the distributed power supply status database, and removing the faulty standby power supply unit from the switchable alternative list.
[0010] Furthermore, the distributed power status database adopts a multi-node distributed architecture, with each storage node deployed on an independent physical server. The nodes synchronize data with each other through a high-speed intranet. Each storage node synchronously stores all power status data. When any storage node fails, the database's built-in load balancing scheduling logic automatically switches subsequent data read and write requests to other normally operating storage nodes. After the failed node is repaired, all updated data during the failure period is automatically synchronized.
[0011] Furthermore, the aforementioned communication components support multi-protocol conversion and are compatible with the communication interfaces of power supply units produced by different manufacturers. The power supply side and the power supply control backend use an encrypted communication protocol for control data transmission. Before each data transmission, two-way identity verification is performed. After successful authentication, a temporary symmetric encryption key is negotiated and generated. All subsequent transmitted power supply control data is encrypted using this temporary key. The temporary key is automatically destroyed after each communication connection is disconnected, and a new key is renegotiated and generated during the next connection.
[0012] Furthermore, the dynamic adjustment of the switching timing is achieved using the following formula: ; in, This refers to the adjustable switching interval between the backup power supply being turned on and the main power supply being turned off. This is the voltage stability coefficient. This is the absolute value of the deviation between the real-time output voltage and the rated voltage of the backup power supply unit. For load adaptability coefficient, This represents the current power supply to the load. This is the working condition correction factor. Rate the current operating condition of the load. This is the load fluctuation correction factor. It is the ratio of the current load power fluctuation value to the rated load power per unit time.
[0013] Furthermore, the sampling period can be dynamically adjusted according to the load operating status. The sampling period is shortened under high dynamic operating conditions and extended under stable operating conditions. The preprocessing of the collected data includes automatically removing abnormal collected data that exceeds the rated parameter range of the corresponding power supply unit, and supplementing the finite number of consecutive missing sampling points by weighted averaging of adjacent valid sampling points. The weighting coefficient is configured according to the time distance between the sampling point and the missing point, and the closer the time distance, the higher the weighting coefficient.
[0014] Furthermore, the entire process control is completed sequentially through four execution stages: communication adaptation, status acquisition, health assessment, and switchover execution. The communication adaptation stage establishes a two-way encrypted control link between each redundant power unit, the status acquisition terminal, and the power supply control backend, completing hash verification and basic data calibration and storage for each redundant power unit. The status acquisition stage collects, preprocesses, and updates the database for multi-dimensional power supply operation parameters. The health assessment stage performs quantitative assessment of the main power unit's health, dynamically adjusts the switchover threshold based on power supply task priorities, and generates redundant power supply switchover commands. The switchover execution stage verifies the switchover commands and performs uninterrupted power supply switchover operations, simultaneously uploading and archiving all switchover process data.
[0015] Furthermore, during the health assessment phase, when quantitatively assessing the health of the main power supply unit, the indicators of real-time power supply operation parameters, historical fault records, number of charge-discharge cycles, and usage duration of the main power supply unit are comprehensively considered. A pre-trained machine learning assessment model is used to output a quantitative health score. The machine learning assessment model is trained using a large amount of power supply operation and fault data accumulated in the past.
[0016] Furthermore, the verification of switching commands during the switching execution phase includes legality verification, timeliness verification, and operating condition matching verification. After each redundant power supply switching operation is completed, the power supply control backend automatically generates a switching operation log and archives the switching-related operating data to the distributed power supply status database.
[0017] Compared with existing technologies, the beneficial effects of this invention are: This invention achieves completely uninterrupted power supply switching by using a slow-start timing control power circuit switching technique. The voltage fluctuation during the switching process is kept at an extremely low level, effectively solving the problems of power supply interruption and voltage spike impact on sensitive components in existing hard switching schemes. This ensures the power supply stability of high-priority tasks and avoids accidents such as crashes and data loss caused by power supply problems.
[0018] This invention significantly improves the accuracy of power health assessment by using multi-dimensional power health evaluation and dynamic adjustment of switching thresholds based on task priority. It effectively reduces the probability of false switching and missed switching, and can flexibly balance power supply reliability and power utilization according to operational needs. It effectively solves the problems of high false switching rate and low power utilization in existing hot standby switching schemes, and can be adapted to different priority operation scenarios, reducing overall operating costs.
[0019] This invention ensures the secure transmission of power supply control commands and operational data by establishing a two-way encrypted control link and deploying a distributed power status database. At the same time, it enables reliable storage and automatic archiving of power supply process data, effectively solving the problems of data being easily tampered with and lacking effective operation and maintenance data support in existing solutions. It can directly provide reliable data support for power supply fault tracing, switching strategy optimization, and power product performance iteration, greatly improving operation and maintenance efficiency.
[0020] This solution is compatible with all industrial and consumer drones equipped with redundant power supply units. It does not require significant modifications to the existing power supply hardware structure and can be directly embedded into the existing drone power supply control system. It has high application value and can effectively improve the power supply reliability and operational efficiency of all types of drones. Attached Figure Description
[0021] Figure 1 This is an overall flowchart of a redundant switching control method for a UAV power supply proposed in this invention; Figure 2 Here is a flowchart of the status acquisition and preprocessing process; Figure 3 A flowchart for health assessment and strategic decision-making; Figure 4 Here is a flowchart of the redundancy switching control execution process; Figure 5 This is a flowchart for the operation, maintenance, and security management of distributed databases. Detailed Implementation
[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this 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. Therefore, they should not be construed as limitations on this invention.
[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.
[0025] Reference Figures 1 to 5 A method for redundant switching control of power supply for unmanned aerial vehicles (UAVs) includes the following steps: Relying on the communication components of the power supply system, a two-way encrypted control link is established between each redundant power supply unit, status acquisition terminal and power supply control backend. The identity, rated parameters and historical operating data of each redundant power supply unit are hash-verified and then synchronously stored in the distributed power supply status database. The power supply basic data calibration is completed before the switching control. The data is verified and compared with the reference parameters of the same type of power supply in the database to avoid parameter deviations that may cause abnormal switching control of subsequent power supply circuits. The status acquisition terminal collects multi-dimensional power supply operation parameters such as output voltage, output current, remaining power, internal temperature, and circuit on / off status of each redundant power supply unit in real time according to a dynamically adjustable preset sampling period. After preliminary preprocessing of the collected data, it transmits the data to the power supply control backend through an encrypted control link and updates the real-time data entries of the distributed power supply status database in a synchronous manner, ensuring the effectiveness and continuity of the power supply operation data in the database. The power supply control backend calls the real-time power supply operation data and historical fault sample data in the distributed power supply status database to quantitatively evaluate the health of the main power supply unit currently in power supply status. The preset switching threshold for health evaluation is dynamically adjusted according to the priority of power supply tasks. Under high-priority power supply tasks, the switching threshold is increased to ensure power supply reliability, while under low-priority power supply tasks, the switching threshold is decreased to improve power utilization. When the health is lower than the preset switching threshold, a redundant power supply switching command adapted to the current operating condition of the load is generated and sent to the switching control execution end. After the switching control execution terminal verifies the switching command, it controls the power supply circuit of the backup power unit to be switched on according to the preset soft start sequence, gradually increases the output duty cycle of the backup power unit, and controls the power supply circuit of the original main power unit to be disconnected after the output voltage stabilizes in the rated range, thus completing the uninterrupted power supply switching. At the same time, all data of the switching process is uploaded to the distributed power status database for archiving and storage.
[0026] This invention also includes a step of real-time operation and maintenance of the distributed power status database, and the database health status is calculated using the following formula: ; in, The health score for the distributed power source state database is a dimensionless parameter. is the weighting coefficient for power supply data integrity, and is a dimensionless constant. The proportion of valid power supply data entries in the database to the total number of data entries is a dimensionless parameter. The weighting coefficient for the response efficiency of power supply data query is a dimensionless constant. The power supply data query response efficiency score is a dimensionless parameter. For power supply data storage security weighting coefficient, which is a dimensionless constant; The power supply data storage security score is a dimensionless parameter. This is a correction factor for abnormal power supply data, and it is a dimensionless constant. The percentage of unarchived abnormal power supply fault data entries in the database is a dimensionless parameter. The three factors satisfy the constraint that their weights sum to 1. The calculated health score directly reflects the overall operating status of the database. It is automatically calculated after each power supply status data synchronization update. When the health score is lower than the preset maintenance threshold, maintenance operations such as data backup, redundant storage node switching, and abnormal entry cleanup are automatically triggered. Data backup adopts a multi-node off-site backup method. After the backup is completed, the consistency between the backup data and the original data is automatically verified. Abnormal entry cleanup releases database storage space, ensuring the storage reliability and retrieval efficiency of power supply status data.
[0027] This invention also includes a step of real-time monitoring of the standby status of each redundant power supply unit. Status query commands are periodically sent to the standby power supply units to obtain their real-time power supply operation parameters. The health of the standby power supply units is pre-assessed to ensure they are always in a ready-to-switch state. When the pre-assessed health of a standby power supply unit is lower than a preset standby threshold, an alarm message is promptly generated and pushed to maintenance personnel. Simultaneously, the standby power supply status identifier in the distributed power supply status database is updated, and the faulty standby power supply unit is removed from the switchable candidate list to prevent subsequent switching commands from specifying the faulty standby power supply unit, which could lead to switching failure.
[0028] In this invention, the distributed power status database adopts a multi-node distributed architecture. Each storage node is deployed on an independent physical server, and the nodes synchronize data through a high-speed intranet. Each storage node synchronously stores all power status data. When any storage node fails, the database's built-in load balancing scheduling logic automatically switches subsequent data read and write requests to other normally operating storage nodes without interrupting the power status data collection and update process. After the faulty node is repaired, all updated data during the fault period is automatically synchronized to ensure data consistency. Fault recovery can be completed without manual intervention.
[0029] In this invention, the aforementioned communication component supports multi-protocol conversion and is compatible with the communication interfaces of power supply units produced by different manufacturers. The power supply side and the power supply control backend use an encrypted communication protocol for control data transmission. Before each data transmission, two-way identity verification is performed. After successful authentication, a temporary symmetric encryption key is negotiated and generated. All subsequent transmitted power supply control data is encrypted using this temporary key. The temporary key is automatically destroyed after each communication connection is disconnected. A new key is renegotiated and generated during the next connection, thus preventing the power supply operation data and switching control commands from being tampered with or stolen, and ensuring the security of power supply control transmission.
[0030] In this invention, the dynamic adjustment of the switching timing is achieved using the following formula: ; in, The time interval to be adjusted between the switching of the backup power supply and the disconnection of the main power supply, in seconds; is the voltage stability coefficient, with dimensions in seconds per volt, and is a constant; The absolute value of the deviation between the real-time output voltage and the rated voltage of the backup power unit, in volts; The load adaptability factor is expressed in seconds per watt and is a constant. The current power supply to the load is expressed in watts. is the working condition correction factor, with the dimension of seconds, and is a constant; The current operating condition of the load is scored, and the score is a dimensionless parameter. is the load fluctuation correction factor, and is a dimensionless constant; This is the ratio of the current load power fluctuation value to the rated load power per unit time, and is a dimensionless parameter. The calculated switching interval can be dynamically adjusted according to the actual operating conditions. The time interval between the backup power supply being turned on and the main power supply being turned off is adjusted according to the real-time operating status of the load to avoid excessive power supply voltage fluctuations during the switching process and to ensure the operational stability of the load equipment.
[0031] In this invention, the sampling period can be dynamically adjusted according to the load operating state. Under high dynamic operating conditions, the sampling period is shortened to increase the power supply data acquisition frequency, while under stable operating conditions, the sampling period is extended to reduce the power consumption of the data acquisition terminal. The preprocessing process of the acquired data includes automatically removing abnormal acquired data that exceeds the rated parameter range of the corresponding power supply unit, and supplementing consecutively missing sampling points by using a weighted average of adjacent valid sampling points. The weighting coefficient is configured according to the time distance between the sampling point and the missing point. The closer the time distance, the higher the weighting coefficient. The amount of data after preprocessing is significantly reduced compared to the amount of original acquired data, reducing the amount of invalid data transmission and the data processing pressure of the control background, and improving the power supply data transmission efficiency.
[0032] In this invention, the entire process control is completed through four execution stages: communication adaptation, status acquisition, health assessment, and switching execution. The communication adaptation stage establishes a bidirectional encrypted control link between each redundant power unit, the status acquisition terminal, and the power supply control backend, completing hash verification and basic data calibration and storage for each redundant power unit. The status acquisition stage collects, preprocesses, and updates the database for multi-dimensional power supply operation parameters. The health assessment stage performs a quantitative assessment of the main power unit's health, dynamically adjusts the switching threshold based on power supply task priorities, and generates redundant power supply switching instructions for the appropriate operating conditions. The switching execution stage verifies the switching instructions and performs uninterrupted power supply switching operations, simultaneously uploading and archiving all switching process data.
[0033] In this invention, when the health assessment stage quantifies the health of the main power supply unit, it comprehensively considers multiple dimensions of indicators such as the real-time power supply operation parameters, historical fault records, charge and discharge cycle count, and usage duration of the main power supply unit. A pre-trained machine learning assessment model is used to output a quantitative health score. The machine learning assessment model is trained using a large amount of accumulated power supply operation and fault data. Compared with the traditional single threshold judgment method, it improves the accuracy of the health assessment results and reduces the occurrence of misjudgments or omissions.
[0034] In the present invention, the verification of the switching instruction in the switching execution phase includes legality verification, timeliness verification and working condition matching degree verification. The legality verification includes instruction source identity verification, instruction format verification and instruction signature verification. The timeliness verification compares the difference between the instruction generation time and the current time, and an instruction whose difference exceeds the preset valid duration is directly determined as an invalid instruction. The working condition matching degree verification compares whether the backup power parameter specified in the instruction is consistent with the actual parameter of the current backup power supply. After each redundant power supply switching operation is completed, the power supply control background automatically generates a complete switching operation log, and archives and stores all the switching trigger conditions, switching timing parameters, operating parameters of each power unit before and after switching, and voltage fluctuation data during the switching process into the distributed power state database. The log storage meets the preset time limit requirements, and can be retrieved and queried at any time, providing reliable data support for subsequent power failure analysis, switching strategy optimization, and power product performance iteration.
[0035] The specific implementation of the present system is further described below through two embodiments: Embodiment 1
[0036] The application scenario of this embodiment is the operation of high-voltage transmission line inspection drones. High-voltage transmission line inspection drones mostly operate in complex field environments such as remote mountainous areas and high-altitude areas, and are equipped with precision inspection loads such as high-definition cameras and infrared temperature measuring devices. No power supply interruption is allowed during the operation process, otherwise it will lead to loss of inspection data, flight control system crash or even drone crash, which is a high-priority power supply task. In this embodiment, the inspection drone is equipped with two sets of redundant power units of the same specification, and the initialization verification of all components is completed before the operation starts.
[0037] In the communication link construction phase, the two-way encrypted control link between each power unit and the power supply control background is constructed relying on the communication components supporting the power supply system. The communication component supports multi-protocol conversion, and can adapt to the communication interfaces of power units produced by different manufacturers. All data transmission between the power supply side and the power supply control background adopts an encrypted communication protocol. Before each data transmission, two-way identity verification is performed first, and a temporary symmetric encryption key is generated through negotiation after the authentication is passed. All subsequent transmitted power supply control data are encrypted with this temporary key, and the temporary key is automatically destroyed after each communication connection is disconnected, and a new key is generated through renegotiation at the next connection, so as to prevent the power supply operation data and switching control instructions from being tampered with or stolen. After the link is constructed, hash verification is performed on the identity identifier, rated parameters and historical operation data of each power unit. After the verification is passed, all basic data are stored in the distributed power state database to complete the basic data calibration.
[0038] The distributed power status database adopts a multi-node distributed architecture, with each storage node deployed on an independent physical server. Nodes synchronize data via a high-speed intranet, and each node synchronously stores all power status data. When any storage node fails, the database's built-in load balancing scheduling logic automatically switches subsequent data read / write requests to other normally operating storage nodes, without interrupting the power status data collection and update process. After the failed node is repaired, all updated data during the failure period is automatically synchronized, ensuring data consistency and enabling fault recovery without manual intervention. After each data update, the distributed power status database automatically calculates a health score. When the health score falls below a preset maintenance threshold, it automatically triggers data backup, redundant storage node switching, and abnormal entry cleanup operations. Data backup uses a multi-node off-site backup method. After backup, the consistency between the backup data and the original data is automatically verified. Abnormal entry cleanup releases database storage space, ensuring the reliability of power status data storage and retrieval efficiency.
[0039] During the inspection operation, the status acquisition terminal dynamically adjusts the sampling cycle according to the drone's operating status. When the drone is in a highly dynamic operating state, such as take-off and landing, obstacle avoidance, or close-to-line inspection, the sampling cycle is shortened to increase the power supply data acquisition frequency. When the drone is in a stable cruise state, the sampling cycle is extended to reduce the power consumption of the data acquisition terminal. The collected power supply operation data is first preprocessed to remove abnormal data that exceeds the rated parameter range of the corresponding power supply unit. For a limited number of consecutively missing sampling points, a weighted average of adjacent valid sampling points is used to fill in the gaps. The weighting coefficient is configured according to the time distance between the sampling point and the missing point; the closer the time distance, the higher the weighting coefficient. The amount of data after preprocessing is significantly reduced compared to the original collected data, reducing the amount of invalid data transmitted and the data processing pressure on the control backend, thus improving the power supply data transmission efficiency. The preprocessed data is synchronously uploaded to the power supply control backend to update the real-time entries in the distributed power supply status database.
[0040] The power supply control backend calls upon real-time data and historical fault sample data to quantitatively assess the health of the main power supply unit. The assessment process comprehensively considers multiple dimensions of indicators, including real-time power supply operation parameters, historical fault records, charge / discharge cycle count, and usage duration. A pre-trained machine learning assessment model is used to output a quantitative health score. This model is trained using a large amount of accumulated power supply operation and fault data, improving the accuracy of health assessment results and reducing false positives or false negatives compared to traditional single-threshold judgment methods. Since inspection tasks are high-priority power supply tasks, the switching threshold for health assessment is set at a relatively high level. The backend synchronously and periodically sends status query commands to the backup power supply unit to obtain its real-time power supply operation parameters and pre-assesses its health. This ensures that the backup power supply unit is always in a ready-to-switch state. When the pre-assessed health of the backup power supply unit falls below the preset standby threshold, an alarm message is promptly generated and pushed to maintenance personnel. Simultaneously, the backup power supply status identifier in the distributed power supply status database is updated, and the faulty backup power supply unit is removed from the switchable candidate list to prevent switching failures caused by specifying the faulty backup power supply unit when generating subsequent switching commands.
[0041] Once the main power supply health falls below a threshold, the backend immediately generates a switching command adapted to the current load conditions and sends it to the switching control execution terminal. Upon receiving the switching command, the switching control execution terminal first performs legality, timeliness, and operating condition matching checks. Legality checks include verifying the source identity of the command, verifying the command format, and verifying the command signature. Timeliness checks compare the difference between the command generation time and the current time; commands with a difference exceeding a preset valid duration are directly deemed invalid. Operating condition matching checks whether the backup power supply parameters specified in the command match the actual parameters of the current backup power supply. After successful verification, the switching interval adapted to the current operating conditions is calculated based on the real-time output voltage deviation of the backup power supply, the current load power, the load operating conditions, and the load power fluctuation value. The time interval between the backup power supply being turned on and the main power supply being turned off is adjusted to avoid excessive voltage fluctuations during the switching process. Subsequently, the power supply circuit of the backup power supply unit to be switched on is controlled according to the soft-start sequence, gradually increasing the output duty cycle of the backup power supply unit. Once the output voltage stabilizes within the rated range, the power supply circuit of the original main power supply unit is disconnected, completing the uninterrupted power supply switching. After each redundant power supply switching operation is completed, the power supply control backend automatically generates a complete switching operation log, which archives and stores the switching trigger conditions, switching timing parameters, operating parameters of each power supply unit before and after the switching, and voltage fluctuation data during the switching process in the distributed power supply status database. The log storage meets the preset period requirements and can be retrieved and queried at any time, providing reliable data support for subsequent power supply fault analysis, switching strategy optimization, and power product performance iteration.
[0042] Table 1: Comparison table of parameters during power switching process of inspection drone
[0043] Table 1 presents the comparison of actual operating parameters of two switching control methods in the operation scenario of high-voltage transmission line inspection drones. The method of the present invention has no power supply interruption at all during the switching process, which can avoid abnormal operation of the inspection payload and flight control system caused by power loss. The maximum voltage fluctuation range is far lower than that of the traditional method, and will not cause impact on the precision electronic components of the drone. The switching success rate is at an extremely high level, which can effectively reduce the risk of crash in the field operation environment. The false switching occurrence rate is only about one tenth of that of the traditional method, which greatly reduces unnecessary power switching operations and extends the overall service life of the power supply. The improvement of power supply data storage reliability also provides stable data support for subsequent power supply fault tracing and switching strategy optimization, fully meets the extremely high requirements of inspection tasks for power supply reliability, and solves the defects of traditional switching methods such as large fluctuation, high false switching rate and insufficient data support.
[0044] Embodiment 2
[0045] The application scenario of this embodiment is the operation of urban last-mile logistics distribution drones. Most urban last-mile logistics distribution drones operate in dense urban airspace, carry distribution parcels, require stable and reliable power supply during operation, and at the same time need to take into account power utilization and control the overall operation cost, which belongs to a medium-priority power supply task. In this embodiment, the logistics drone is equipped with three sets of redundant power supply units of the same specification, and initialization verification of all components is completed before the operation starts.
[0046] In the communication link construction stage, the two-way encrypted control link between each power supply unit and the power supply control background is constructed relying on the communication components supporting the power supply system. The communication component supports multi-protocol conversion, and can adapt to the communication interfaces of power supply units produced by different manufacturers. All data transmission between the power supply side and the power supply control background adopts an encrypted communication protocol. Before each data transmission, two-way identity verification is performed first, and a temporary symmetric encryption key is negotiated and generated after the authentication is passed. All subsequently transmitted power supply control data are encrypted with this temporary key, and the temporary key is automatically destroyed after each communication connection is disconnected, and a new key is regenerated through negotiation at the next connection, so as to prevent the power supply operation data and switching control instructions from being tampered with or stolen. After the link is constructed, hash verification is performed on the identity identifier, rated parameters and historical operation data of each power supply unit. After the verification is passed, all basic data are stored in the distributed power supply state database to complete basic data calibration.
[0047] The distributed power status database adopts a multi-node distributed architecture, with each storage node deployed on an independent physical server. Nodes synchronize data via a high-speed intranet, and each node synchronously stores all power status data. When any storage node fails, the database's built-in load balancing scheduling logic automatically switches subsequent data read / write requests to other normally operating storage nodes, without interrupting the power status data collection and update process. After the failed node is repaired, all updated data during the failure period is automatically synchronized, ensuring data consistency and enabling fault recovery without manual intervention. After each data update, the distributed power status database automatically calculates a health score. When the health score falls below a preset maintenance threshold, it automatically triggers data backup, redundant storage node switching, and abnormal entry cleanup operations. Data backup uses a multi-node off-site backup method. After backup, the consistency between the backup data and the original data is automatically verified. Abnormal entry cleanup releases database storage space, ensuring the reliability of power status data storage and retrieval efficiency.
[0048] During delivery operations, the status acquisition terminal dynamically adjusts the sampling cycle based on the drone's operational status. When the drone is in highly dynamic operational states such as takeoff and landing, obstacle avoidance, and package delivery, the sampling cycle is shortened to increase the frequency of power supply data acquisition. When the drone is in a straight-line cruise state, the sampling cycle is extended to reduce the power consumption of the data acquisition terminal. The collected power supply operation data is first preprocessed to remove abnormal data exceeding the rated parameter range of the corresponding power supply unit. For a limited number of consecutively missing sampling points, a weighted average of adjacent valid sampling points is used to fill in the gaps. The weighting coefficient is configured based on the time distance between the sampling point and the missing point; the closer the time distance, the higher the weighting coefficient. The amount of data after preprocessing is significantly reduced compared to the original collected data, reducing the amount of invalid data transmitted and the data processing pressure on the control backend, thus improving the efficiency of power supply data transmission. The preprocessed data is synchronously uploaded to the power supply control backend to update the real-time entries in the distributed power supply status database.
[0049] The power supply control backend calls upon real-time data and historical fault sample data to quantitatively assess the health of the main power supply unit. The assessment process comprehensively considers multiple dimensions of indicators, including real-time power supply operation parameters, historical fault records, charge / discharge cycle count, and usage duration. A pre-trained machine learning assessment model outputs a quantitative health score. This model is trained using a large amount of accumulated power supply operation and fault data, improving the accuracy of health assessment results and reducing false positives or false negatives compared to traditional single-threshold judgment methods. Since logistics and delivery tasks are considered medium-priority power supply tasks, the switching threshold for health assessment is set at a low level to maximize power utilization and reduce operating costs while ensuring power supply safety. The backend synchronously and periodically sends status query commands to the two sets of backup power units to obtain the real-time power supply operation parameters of each backup power unit, and performs a pre-assessment of the health of each backup power unit to ensure that the backup power units are always in a standby state that can be switched on normally. When the pre-assessed health of any backup power unit is lower than the preset standby threshold, an alarm message is generated in a timely manner and pushed to the operation and maintenance personnel. At the same time, the backup power status identifier in the distributed power status database is updated, and the faulty backup power unit is removed from the switchable alternative list to avoid the failure of switching due to specifying the faulty backup power unit when generating switching commands in the future.
[0050] Once the main power supply health falls below a threshold, the backend immediately generates a switching command adapted to the current load conditions and sends it to the switching control execution terminal. Upon receiving the switching command, the switching control execution terminal first performs legality, timeliness, and operating condition matching checks. Legality checks include verifying the source identity of the command, verifying the command format, and verifying the command signature. Timeliness checks compare the difference between the command generation time and the current time; commands with a difference exceeding a preset valid duration are directly deemed invalid. Operating condition matching checks whether the backup power supply parameters specified in the command match the actual parameters of the current backup power supply. After successful verification, the switching interval adapted to the current operating conditions is calculated based on the real-time output voltage deviation of the backup power supply, the current load power, the load operating conditions, and the load power fluctuation value. The time interval between the backup power supply being turned on and the main power supply being turned off is adjusted to avoid excessive voltage fluctuations during the switching process. Subsequently, the power supply circuit of the backup power supply unit to be switched on is controlled according to the soft-start sequence, gradually increasing the output duty cycle of the backup power supply unit. Once the output voltage stabilizes within the rated range, the power supply circuit of the original main power supply unit is disconnected, completing the uninterrupted power supply switching. After each redundant power supply switching operation is completed, the power supply control backend automatically generates a complete switching operation log, which archives and stores the switching trigger conditions, switching timing parameters, operating parameters of each power supply unit before and after the switching, and voltage fluctuation data during the switching process in the distributed power supply status database. The log storage meets the preset period requirements and can be retrieved and queried at any time, providing reliable data support for subsequent power supply fault analysis, switching strategy optimization, and power product performance iteration.
[0051] Table 2: Comparison of Power Supply Performance of Logistics Drones
[0052] Table 2 presents a comparison of the actual operational efficiency of the two switching control methods in urban last-mile logistics delivery drone operation scenarios. The method of this invention, by dynamically adjusting the switching threshold, significantly improves the average utilization rate of a single power supply while ensuring power supply safety, extends the flight time of a single flight, effectively reduces the energy consumption cost per delivery order, and maintains an extremely low switching-related failure rate. This significantly reduces losses such as crashes and lost parts due to power supply failures during operation. The power supply maintenance cost accounts for only about one-third of that of the traditional method, reducing overall operating expenses. The improved response speed for power supply data retrieval also facilitates rapid troubleshooting by maintenance personnel, shortening the drone's maintenance turnaround cycle. It fully adapts to the low-cost, high-turnaround operational needs of logistics delivery scenarios, solving the shortcomings of traditional switching methods such as low power utilization, high operating costs, and insufficient maintenance efficiency.
[0053] Reference Figure 1This diagram illustrates the overall workflow and progressive relationship of control phases in the redundant switching control of UAV power supplies. The entire control process consists of four interconnected execution phases. In the control preparation phase, the system first enters the communication adaptation stage, establishing a bidirectional encrypted control link between each redundant power unit, status acquisition terminal, and control backend using supporting communication components, and performing hash verification and calibration storage on the basic data. Subsequently, the system enters the status acquisition phase, where the terminal periodically acquires multi-dimensional parameters and updates them to the database. In the health assessment phase, the backend uses a machine learning model to quantitatively assess the main power supply and dynamically adjusts the switching threshold based on the current task priority. Once the main power supply health triggers the threshold, the system immediately enters the switching execution phase. After completing multiple verifications, it controls the backup power supply to slowly start and disconnects the original power supply, ultimately achieving uninterrupted power supply switching and full data archiving.
[0054] Reference Figure 2 This diagram illustrates the efficient implementation process of power supply status data acquisition, dynamic adjustment, and preprocessing at the data acquisition terminal. This process aims to ensure the validity and continuity of power supply operation data within the distributed database, while reducing the amount of invalid data transmission. The control system first dynamically adjusts the sampling period of the status acquisition terminal based on the actual operating status of the UAV: automatically shortening the sampling period to increase the data acquisition frequency under high-dynamic operating conditions, and extending the sampling period to reduce power consumption at the acquisition terminal under stable operating conditions. Within the set period, the acquisition terminal acquires multi-dimensional parameters in real time, including output voltage, output current, remaining power, internal temperature, and circuit continuity status of each redundant power supply. Before data upload, the terminal automatically performs preprocessing, removing bad data exceeding the rated range of the corresponding power supply, and performing weighted averaging of adjacent valid points to complete the finite number of consecutive missing sampling points based on time distance, finally uploading the simplified valid data synchronously.
[0055] Reference Figure 3 This diagram illustrates the intelligent control process of the system in quantifying the health of the current main power unit and making switching decisions. By introducing a dynamic threshold mechanism, the system can perfectly balance the safety requirements and economic benefits of the UAV under different flight missions. The power supply control backend first retrieves multi-dimensional indicators such as real-time power supply data, historical fault records, charge / discharge cycle count, and usage duration of the main power unit from the distributed power status database. Subsequently, these indicators are input into a pre-trained machine learning evaluation model, which outputs a quantitative health score that accurately reflects the power supply status through correlation analysis. Based on this, the backend dynamically adjusts the preset switching threshold according to the current power supply task priority: the threshold is increased when performing high-priority power supply tasks to ensure ultimate reliability, and the threshold is decreased under low-priority tasks to fully utilize the power supply. When the score falls below the current dynamic threshold, the system immediately issues a switching command adapted to the current load condition.
[0056] Reference Figure 4 This diagram details how the switching control actuator, upon receiving an action command, achieves seamless and safe switching of the power supply circuit through rigorous logic verification and precise timing control. Upon receiving the switching command, the actuator first initiates a highly secure three-dimensional verification process, including verification of the command's source, format, and signature; timeliness verification by comparing the generation time difference; and verification of the matching degree of the specified backup power supply parameters. After passing the verification, the system dynamically calculates the optimal switching interval between the backup power supply's activation and the main power supply's deactivation based on the load's current power fluctuations and operating condition score. During the execution phase, the actuator strictly follows the preset soft-start sequence, first activating the backup power supply circuit and gradually increasing its output duty cycle. Once the backup power supply's output voltage stabilizes within the rated range, the original main power supply circuit is then disconnected, completing a lossless switching while avoiding significant fluctuations in the supply voltage, and finally generating and archiving logs.
[0057] Reference Figure 5 This diagram illustrates the network security defense mechanisms, multi-node architecture design, and automated intelligent operation and maintenance processes of the distributed power supply status database during operation, aiming to ensure the reliability of power supply status data storage and the efficiency of retrieval. At the security level, the power supply side and the control backend perform two-way identity verification and negotiate to generate a temporary symmetric encryption key before each communication, which is automatically destroyed upon disconnection to ensure transmission security. In terms of storage architecture, the database adopts a distributed architecture composed of multiple physical server nodes. Nodes synchronize all data through a high-speed internal network, and in the event of a node failure, built-in load balancing scheduling logic enables automatic and seamless switching of read and write requests. At the system operation and maintenance level, after each status data synchronization update, the database automatically performs weighted calculations on data integrity, query response efficiency, storage security, and the proportion of unarchived faulty data. When the health score falls below a preset operation and maintenance threshold, the system automatically triggers multi-node off-site backup and abnormal entry cleanup, achieving fault recovery and space release without manual intervention.
[0058] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A method for redundant switching control of power supply for unmanned aerial vehicles (UAVs), characterized in that, Includes the following steps: Relying on the communication components of the power supply system, a two-way encrypted control link is established between each redundant power supply unit, status acquisition terminal and power supply control backend. The identity, rated parameters and historical operating data of each redundant power supply unit are hash-verified and stored in the distributed power supply status database. Based on the reference parameters of the same type of power supply in the database, the basic power supply data calibration is completed. The status acquisition terminal collects multi-dimensional power supply operation parameters of each redundant power supply unit in real time according to a dynamically adjustable preset sampling period. After preliminary preprocessing of the collected data, it transmits the data to the power supply control backend through an encrypted control link and updates the real-time data entries of the distributed power supply status database synchronously. The power supply control backend retrieves real-time operating data and historical fault sample data from the distributed power supply status database, quantitatively assesses the health of the main power supply unit, adjusts the health switching threshold according to the power supply task priority, and sends a redundant power supply switching command to the switching control execution terminal when the health is lower than the switching threshold. After the switching control execution terminal verifies the switching command, it controls the power supply circuit of the backup power unit to be switched on to be turned on according to the preset soft start sequence. After the output voltage stabilizes in the rated range, it controls the power supply circuit of the original main power unit to be disconnected, completing the uninterrupted power supply switching. At the same time, all data of the switching process is uploaded to the distributed power status database for archiving and storage.
2. The method for redundant switching control of UAV power supply according to claim 1, characterized in that, It also includes the step of real-time operation and maintenance of the distributed power status database. The database health status is calculated using the following formula: ; in, A health score is assigned to the distributed power state database. For power supply data integrity weighting coefficient, This refers to the proportion of valid power supply data entries in the database out of the total number of data entries. The power supply data query response efficiency weighting coefficient. The power supply data query response efficiency is scored. For power supply data storage security weighting coefficient, Assess the security of power supply data storage. Correction factor for abnormal power supply data. This represents the percentage of unarchived abnormal power supply fault data entries in the database. The three satisfy the constraint that the sum of their weights is 1.
3. The method for redundant switching control of UAV power supply according to claim 1, characterized in that, It also includes steps for real-time monitoring of the standby status of each redundant power supply unit, periodically issuing status query commands to the standby power supply unit, collecting real-time power supply operation parameters, conducting a pre-assessment of the health of the standby power supply unit based on the collected parameters, generating alarm information when the pre-assessed health is lower than the standby threshold, synchronously updating the standby power supply status identifier in the distributed power supply status database, and removing the faulty standby power supply unit from the switchable alternative list.
4. The method for redundant switching control of UAV power supply according to claim 2, characterized in that, The distributed power status database adopts a multi-node distributed architecture. Each storage node is deployed on an independent physical server, and the nodes synchronize data through a high-speed intranet. Each storage node synchronously stores all power status data. When any storage node fails, the database's built-in load balancing scheduling logic automatically switches subsequent data read and write requests to other normally operating storage nodes. After the failed node is repaired, all updated data during the failure period is automatically synchronized.
5. The method for redundant switching control of UAV power supply according to claim 1, characterized in that, The aforementioned communication component supports multi-protocol conversion and is compatible with the communication interfaces of power supply units produced by different manufacturers. The power supply side and the power supply control backend use an encrypted communication protocol for control data transmission. Before each data transmission, two-way identity verification is performed. After successful authentication, a temporary symmetric encryption key is negotiated and generated. All subsequent transmitted power supply control data is encrypted using this temporary key. The temporary key is automatically destroyed after each communication connection is disconnected, and a new key is renegotiated and generated during the next connection.
6. The method for redundant switching control of UAV power supply according to claim 3, characterized in that, The dynamic adjustment of the switching timing is performed using the following formula: ; in, This refers to the adjustable switching interval between the backup power supply being turned on and the main power supply being turned off. This is the voltage stability coefficient. This is the absolute value of the deviation between the real-time output voltage and the rated voltage of the backup power supply unit. For load adaptability coefficient, This represents the current power supply to the load. This is the working condition correction factor. Rate the current operating condition of the load. This is the load fluctuation correction factor. It is the ratio of the current load power fluctuation value to the rated load power per unit time.
7. The method for redundant switching control of UAV power supply according to claim 1, characterized in that, The sampling period can be dynamically adjusted according to the load operating status. The sampling period is shortened under high dynamic operating conditions and extended under stable operating conditions. The preprocessing of the collected data includes automatically removing abnormal collected data that exceeds the rated parameter range of the corresponding power supply unit, and supplementing the finite number of consecutive missing sampling points by weighted averaging of adjacent valid sampling points. The weighting coefficient is configured according to the time distance between the sampling point and the missing point, and the closer the time distance, the higher the weighting coefficient.
8. The method for redundant switching control of UAV power supply according to claim 1, characterized in that, The entire process control is completed through four execution phases: communication adaptation, status acquisition, health assessment, and switchover execution. The communication adaptation phase establishes a two-way encrypted control link between each redundant power unit, the status acquisition terminal, and the power supply control backend, and completes hash verification and basic data calibration and storage for each redundant power unit. The status acquisition phase completes the acquisition, preprocessing, and database update of multi-dimensional power supply operation parameters. The health assessment phase completes the quantitative assessment of the main power unit's health, dynamically adjusts the switchover threshold based on the power supply task priority, and generates redundant power supply switchover instructions. The switchover execution phase completes the switchover instruction verification and uninterrupted power supply switchover operation, and synchronously uploads and archives all switchover process data.
9. A method for redundant switching control of a UAV power supply according to claim 8, characterized in that, During the health assessment phase, when quantitatively assessing the health of the main power supply unit, the indicators of real-time power supply operation parameters, historical fault records, charge and discharge cycle count, and usage duration of the main power supply unit are comprehensively considered. A pre-trained machine learning assessment model is used to output a quantitative health score. The machine learning assessment model is trained using a large amount of power supply operation and fault data accumulated in the past.
10. A method for redundant switching control of a UAV power supply according to claim 8, characterized in that, The verification of switching commands during the switching execution phase includes legality verification, timeliness verification, and operating condition matching verification. After each redundant power supply switching operation is completed, the power supply control backend automatically generates a switching operation log and archives the switching-related operating data to the distributed power supply status database.
