A multi-rotor unmanned aerial vehicle battery selection optimization and verification method

Abstract

The application belongs to the field of unmanned aerial vehicles, and discloses a multi-rotor unmanned aerial vehicle battery type selection optimization and verification method, which comprises the following steps: S1, acquiring the rated working voltage range of all main battery power taking equipment of the unmanned aerial vehicle and calculating the system safety voltage window; S2, screening the feasible battery platform based on the system safety voltage window, calculating the feasible string number and constructing the candidate battery scheme pool; S3, determining the test voltage domain covering all candidate batteries in the candidate scheme pool, acquiring the corresponding data of voltage, throttle and tension through power test, and establishing a parameterized mathematical model; S4, performing multi-constraint verification and iterative optimization on the candidate scheme based on the parameterized mathematical model to determine the optimal battery type selection scheme; and S5, carrying out flight verification according to the optimal battery type selection scheme, recording the measured data and correcting the model, and archiving the full data of this type selection to construct a design knowledge base.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicles, specifically to a method for selecting, optimizing, and verifying batteries for multi-rotor drones. Background Technology

[0002] The flight performance, endurance, operational reliability, and safety of multi-rotor drones are highly dependent on the compatibility between the power battery system and the overall drone. As the core power source of drones, the voltage compatibility, energy density, power characteristics, weight, and cost of the power battery directly determine the drone's operational efficiency, mission radius, and operational risks.

[0003] The core technical challenge in battery selection within the industry today is the inability to effectively resolve the cyclical dependency between weight and performance. Specifically, battery weight directly determines the total weight of the drone, which in turn determines the thrust required for flight. Thrust requirements, in turn, impose specific demands on battery capacity (energy) and discharge rate (power), forming a closed-loop dependency relationship between battery weight, total drone weight, thrust requirements, battery performance requirements, and battery weight. Traditional selection methods often rely on engineer experience or simple calculations, failing to establish a mathematical solution to this cyclical dependency. This results in either choosing large-capacity, high-rate batteries to meet thrust requirements, leading to excessive drone weight and reduced flight endurance; or choosing small-capacity batteries to control weight, resulting in insufficient power redundancy and an inability to handle high loads or emergency conditions. This contradiction often leads to inconsistencies between weight, power, and energy in design outcomes, a core reason for drones failing to achieve expected flight times, insufficient power reserves, and even accidental crashes.

[0004] Meanwhile, existing battery selection methods lack precise modeling of power performance covering the full discharge range and do not form a closed-loop mechanism for design, verification, and correction. This further exacerbates the difficulty in resolving the cyclical dependency contradiction, making it difficult to achieve a globally optimal match between battery safety, performance, weight, and cost. Therefore, there is an urgent need for a battery selection optimization and verification method that can systematically resolve this cyclical dependency while taking into account both precise modeling and closed-loop verification. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing and verifying the selection of batteries for multi-rotor unmanned aerial vehicles (UAVs) to solve the problem of the cyclical dependence contradiction between weight and performance in battery selection.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for selecting, optimizing, and verifying batteries for multi-rotor unmanned aerial vehicles (UAVs) includes the following steps: S1. Obtain the rated operating voltage range of all devices on the drone that draw power from the main battery, and calculate the system's safe voltage window; S2. Based on the system's safe voltage window, select feasible cell platforms, calculate the number of feasible strings, and construct a pool of candidate battery solutions. S3. Determine the full range of test voltages for all candidate batteries in the candidate solution pool, obtain corresponding data on voltage, throttle and tension, and current through power testing, and establish a parameterized mathematical model. S4. Based on the parameterized mathematical model, the candidate schemes are verified and iteratively optimized through multiple constraints to determine the optimal battery selection scheme; S5. Conduct flight verification based on the optimal battery selection scheme, record the measured data and correct the model, and archive all the data from this selection to build a design knowledge base.

[0007] Preferably, step S1 specifically includes: S11. Obtain and record the rated operating voltage range of all electrical devices on the drone that draw power from the main battery cart. ,in, For device indexing, This refers to the minimum rated operating voltage of the equipment. This is the maximum rated operating voltage of the equipment.

[0008] S12, retrieve all devices The maximum value is the system's minimum operating voltage. All equipment The minimum value is the system's highest operating voltage. The system safety voltage window is obtained as follows: Furthermore, the output voltage of the candidate battery pack falls within the system's safe voltage window throughout the entire charge-discharge cycle.

[0009] Preferably, step S2 specifically includes: S21. Based on the system's safe voltage window and combined with project requirements, select feasible cell platforms and record their key parameters; S22. For each selected cell platform, calculate its feasible series count range and the single-cell operating voltage range. The lowest number of series in the battery cells. , the highest number of strings The number of feasible strings is to satisfy positive integers, in, For the index of the battery cell platform, For the first The minimum operating voltage of a single cell in this type of battery is... For the first The maximum operating voltage of a single cell in this type of battery is It is a rounding function. This is the floor function. For the first The minimum feasible number of series connections for this type of battery cell. For the first The maximum feasible number of series connections for this type of battery cell. This represents the actual number of battery cells connected in series. S23. Within the feasible number of strings for each cell platform, the highest number of strings is selected as the matching number of strings for that cell platform. Commercially available discrete capacity specifications that match each cell platform and the matching string number combination are screened and relevant information is recorded to form a structured candidate solution pool.

[0010] Preferably, step S3 specifically includes: S31. Traverse all candidate batteries in the candidate solution pool and extract the lowest cutoff voltage value among all batteries, denoted as... The highest fully charged voltage value is denoted as Then the test voltage over the entire range is ; S32. Select no fewer than 5 key voltage points within the entire test voltage range, and the key voltage points must include... and The two endpoints and the remaining intermediate points are selected according to the principle of uniform distribution or distribution near typical voltage values; S33. Construct a dedicated power test platform that includes a programmable DC power supply, the power unit under test, a measurement system, and a control and recording system; S34. Adjust the output voltage of the programmable DC power supply to any key voltage point and keep it stable. Increase the throttle from 0% linearly in fixed steps to the maximum value allowed by the ESC. After the system output stabilizes at each throttle point, record the steady-state tension and bus current data at that throttle point through the measurement system. Complete the test of all key voltage points in sequence and summarize to obtain the full power test data of the UAV. S35. Based on the full power test data of the UAV, a parameterized mathematical model describing the relationship between voltage, throttle and tension, and the relationship between voltage, throttle and current is established by using a bivariate polynomial surface fitting method or a piecewise bilinear interpolation method.

[0011] Preferably, the specific process of the bivariate polynomial surface fitting method in step S35 is as follows: assuming that the tension, current, voltage, and throttle are bivariate polynomial relationships, the full test data of the UAV power is substituted into the system of overdetermined equations, and the coefficients are solved using the least squares method to obtain the tension model and the current model.

[0012]

[0013] in, Input voltage for the power system, This represents the percentage of throttle output. Let the highest polynomial order of the voltage term be . Let the highest polynomial order of the throttle term be . The polynomial order variable is the voltage term. Let be the polynomial order variable of the throttle term. The coefficients to be determined for the bivariate polynomial of the tension model are... The coefficients to be determined for the bivariate polynomial of the current model are denoted as . For the order variable of the voltage term in the current model, Let be the order variable of the throttle term in the current model. The thrust values ​​of the power system under different voltages and throttle positions. These represent the powertrain bus current values ​​under different voltages and throttle positions.

[0014] Preferably, the specific process of the piecewise bilinear interpolation method in step S35 is as follows: construct a two-dimensional grid of voltage and throttle based on the full power test data of the UAV, and determine the voltage, throttle and tension of the four adjacent nodes of any given voltage and throttle grid cell; First, calculate the normalized coordinates for the voltage and throttle dimensions:

[0015] Then, the estimated pulling force or current value corresponding to any given voltage and throttle is calculated using the bilinear interpolation formula:

[0016] in, For the first Voltage values ​​of each grid node, For the first Voltage values ​​of each grid node, For the first Throttle value of each grid node For the first Throttle value of each grid node For the voltage dimension, normalized coordinates, For the normalized coordinates of the throttle dimension, The interpolated value represents the tension or current at the target point. For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current.

[0017] Preferably, step S4 specifically includes: S41. Select either discrete specification screening mode or continuous capacity optimization mode according to project requirements. The discrete specification screening mode is suitable for commercially available battery selection scenarios, while the continuous capacity optimization mode is suitable for customized cell R&D scenarios. S42. Based on the parameterized mathematical model established in S3, perform multi-constraint verification and iterative optimization on each candidate scheme in the candidate scheme pool. In the discrete specification screening mode, solve the cyclic dependency problem between battery weight and total aircraft weight through weight closed-loop iteration. In the continuous capacity optimization mode, search for battery capacity that satisfies all constraints with the goal of minimizing total aircraft weight. S43. For candidate schemes that have passed multi-constraint verification and iterative optimization, a comprehensive score is calculated by weighting multiple dimensions such as voltage platform, range, weight efficiency, cost, power margin, and availability. The candidate scheme with the highest comprehensive score is selected as the optimal battery selection scheme.

[0018] Preferably, the multi-constraint verification and iterative optimization of the discrete specification screening mode in S42 includes the following steps: A1. Determine the design parameters of the UAV; A2. Calculate the maximum permissible battery weight, and retain specifications where the actual battery weight is less than the maximum permissible battery weight;

[0019] A3. Estimate the center value of battery capacity using empirical formulas, set the battery capacity range, and generate a preliminary candidate specification list within the battery capacity range:

[0020] in, This is the center value of the battery capacity. The sum of the aircraft's empty weight and maximum mission load. The empirical coefficient for hovering power density is taken as 100-150 W / kg. This is the battery's nominal voltage. The target hovering time, For battery discharge efficiency, For continuous power safety factor; A4. Perform closed-loop iteration of weight for each specification in the initial candidate specification list until the weight converges, and verify that the battery meets the power, energy and power constraints under the cutoff voltage. If the verification is successful, calculate the accurate weight of the battery and converge. A5. Calculate a comprehensive score for the verified specifications and select the specification with the highest comprehensive score as the optimal solution.

[0021] Preferably, the multi-constraint verification and iterative optimization of the continuous capacity optimization mode in S42 includes the following steps: B1. Using the customized nominal battery capacity as the continuous optimization variable, minimizing the total weight of the aircraft as the objective function, and setting energy, continuous power, peak power and power redundancy as the core constraints; B2. The initial search range is determined by combining the upper limit of the hard constraint based on the maximum takeoff weight, the empirical estimation range based on the hovering time, and the rapid lower limit estimation based on the peak power requirement. B3. Use the bisection method to iteratively search for the minimum capacity that satisfies the constraints within the initial search range; B4. The theoretical optimal battery capacity is factored into manufacturing tolerances and rounded to the standard manufacturing steps of the cell manufacturer to obtain the final customized capacity, and a customized cell technical specification is generated.

[0022]

[0023] in, The nominal capacity is determined by taking into account manufacturing tolerances, with manufacturing tolerances set at ±5%. The final customized capacity after rounding is represented by the standard manufacturing increment from the cell manufacturer.

[0024] Preferably, step S5 specifically includes: S51. The optimal battery selection scheme is used to conduct actual flight tests. The safety boundary verification flight and the typical mission profile flight are completed in sequence. The safety boundary verification flight is to conduct a full-throttle vertical climb test in a safe airspace when the battery charge is about 30%. The typical mission profile flight needs to fully execute the simulated mission process of takeoff, climb, cruise, hovering, maneuvering and landing. S52. During the entire flight test, the measured data stream of battery voltage, current and time is recorded at a high frequency. S53. Compare the measured data stream with the predicted data stream of the parameterized mathematical model, and calculate the average current correction coefficient:

[0025] in, This is the current averaging correction factor. The average bus current during the critical phase of flight testing. The average bus current for the corresponding critical stage predicted by the model; S54. Feed the average current correction coefficient back to the multi-constraint verification and iterative optimization process in step S4 to adjust the parameterized mathematical model. S55. Archive all data from this selection process, including input parameters, process data, model parameters, correction coefficients, and flight verification reports, to build a design knowledge base.

[0026] By adopting the above technical solution, the present invention has the following advantages compared with the prior art: 1. This invention provides a method for battery selection optimization and verification for multi-rotor UAVs. By testing and modeling the full voltage power envelope, the power system performance within the full discharge range of the battery is accurately characterized. Combined with a weight-aware iterative optimization algorithm, the cyclical dependence of weight, thrust, and battery performance is broken. Then, through multi-dimensional quantitative scoring, the global optimal balance of performance, weight, cost, and availability is achieved, avoiding performance redundancy or design shortcomings.

[0027] 2. This invention provides a method for selecting, optimizing, and verifying batteries for multi-rotor UAVs. It transforms the experience-based selection process into a standardized and streamlined engineering approach. Through steps such as candidate pool construction, rapid initial screening, and precise iterative verification, it significantly shortens the design cycle and reduces the cost of physical trial and error. At the same time, the quantitative decision-making system and complete process data archiving ensure that the selection results are traceable and reproducible, reducing reliance on personal experience.

[0028] 3. This invention provides a method for selecting, optimizing, and verifying batteries for multi-rotor UAVs. By defining the boundaries of system electrical parameters, the safe voltage window of all electrical equipment in the UAV is clearly defined, ensuring voltage compatibility of candidate batteries throughout the charging and discharging process, and eliminating risks such as overvoltage damage and undervoltage failure from the source. At the same time, based on power redundancy verification under the cutoff voltage, sufficient power is ensured under harsh operating conditions such as low power and high load, significantly reducing the probability of crash.

[0029] 4. This invention provides a method for selecting, optimizing, and verifying batteries for multi-rotor drones. It offers dual modes of discrete specification screening and continuous capacity optimization, supporting rapid selection of commercially available standard batteries while also providing precise customized cell specifications for special drones or those with extreme performance requirements, thus expanding the applicability of the selection method.

[0030] 5. This invention provides a method for optimizing and verifying battery selection for multi-rotor UAVs. Through flight testing and a closed-loop correction mechanism, it calculates correction coefficients using measured data and feeds back to optimize the selection model. Simultaneously, it constructs a design knowledge base containing input parameters, process data, model parameters, and verification reports, providing experience for subsequent similar projects. This enables the selection method to self-improve and continuously upgrade, constantly improving design accuracy and reliability. Attached Figure Description

[0031] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Example

[0033] Please refer to Figure 1 As shown, this invention discloses a method for battery selection optimization and verification for multi-rotor unmanned aerial vehicles (UAVs), comprising the following steps: S1. Obtain the rated operating voltage range of all devices on the UAV that draw power from the main battery, including but not limited to: flight controller, electronic speed controller and motors, servo motors, image data radio link, gimbal and camera, onboard computer, various mission payloads (such as mapping radar, delivery mechanism), etc., and calculate the system's safe voltage window. S2. Based on the system safety voltage window and combined with project requirements, select feasible cell platforms, calculate the feasible string number range for each cell platform, and prioritize the highest string number within this range. Select commercially available discrete capacity specifications that match each cell platform and string number combination to form a structured candidate pool. S3. Determine the full range of test voltages for all candidate batteries in the candidate solution pool. Select key voltage points including the two endpoints within the full range of test voltages. Build a dedicated power test platform and adjust the throttle at each key voltage point in fixed steps. Simultaneously record the full power test data of the UAV at each throttle point. Based on the full power test data of the UAV, establish a parameterized mathematical model of voltage, throttle, tension, and current using a bivariate polynomial surface fitting method or a piecewise bilinear interpolation method. S4. Based on the parameterized mathematical model, the discrete specification screening mode or the continuous capacity optimization mode is used to perform multi-constraint verification and iterative optimization on each candidate scheme in the candidate scheme pool, and the optimal battery selection scheme is determined from the verified candidate schemes through multi-objective quantitative scoring. S5. Conduct safety boundary verification flights and typical mission profile flights based on the optimal battery selection scheme, record the measured data stream throughout the process, compare it with the predicted data stream, calculate the average current correction coefficient, and feed it back to the verification and optimization process in S4. At the same time, archive all the data from this selection to build a design knowledge base.

[0034] Step S1 is as follows: S11. Obtain and record the rated operating voltage range of all electrical devices on the drone that draw power from the main battery cart. ,in, For device indexing, This refers to the minimum rated operating voltage of the equipment. This is the maximum rated operating voltage of the equipment.

[0035] S12, retrieve all devices The maximum value is the system's minimum operating voltage. This ensures that all devices can still function normally even at the lowest voltage. The minimum value is the system's highest operating voltage. This prevents any equipment from being damaged by overvoltage at its highest point, thus achieving a safe voltage window for the system. Furthermore, the output voltage of the candidate battery pack falls within the system's safe voltage window throughout the entire charge-discharge cycle.

[0036] Step S2 is as follows: S21. Based on the system safety voltage window and combined with project requirements (energy density, power density, rate characteristics, cost, high and low temperature performance, cycle life, etc.), select feasible cell platforms and record their key parameters. Individual unit operating voltage range: Nominal voltage: (Typically 3.6V, 3.7V, 3.75V or 3.2V) Typical mass energy density: (Unit: Wh / kg) Typical continuous discharge rate:

[0037] Typical peak discharge rate (e.g., 10 seconds):

[0038] S22. For each selected cell platform, calculate its feasible series count range and the single-cell operating voltage range. The lowest number of series in the battery cells. , the highest number of strings The number of feasible strings is to satisfy positive integers, in, For the index of the battery cell platform, For the first The minimum operating voltage of a single cell in this type of battery is... For the first The maximum operating voltage of a single cell in this type of battery is It is a rounding function. This is the floor function. For the first The minimum feasible number of series connections for this type of battery cell. For the first The maximum feasible number of series connections for this type of battery cell. This represents the actual number of battery cells connected in series. S23. Within the feasible string count range for each cell platform, the highest string count is prioritized as the matching string count for that cell platform. For example, if the calculated feasible string count for a certain cell is 6S-12S, then the 12S solution is prioritized for inclusion in the solution pool. Commercially available discrete capacity specifications that match each cell platform and matching string count combination are screened and relevant information is recorded to form a structured candidate solution pool, as shown in Table 1 below: Table 1. Structured Scheme Database Step S3 is as follows: S31. Traverse all candidate batteries in the candidate solution pool and extract the lowest cutoff voltage value among all batteries, denoted as... The highest fully charged voltage value is denoted as Then the test voltage over the entire range is ; S32. Select no fewer than 5 key voltage points within the entire test voltage range, and the key voltage points must include and The two endpoints and the remaining intermediate points are selected according to the principle of uniform distribution or distribution near typical voltage values; S33. Construct a dedicated power test platform that includes a programmable DC power supply, a power unit under test, a measurement system, and a control and recording system. The programmable DC power supply is used to accurately simulate and output various test voltage points. The power unit under test includes a motor, an ESC, and a propeller. The measurement system includes a high-precision thrust meter, a current sensor, and a data acquisition card. The control and recording system is used to send throttle signals and synchronously record all sensor data.

[0039] S34. Adjust the output voltage of the programmable DC power supply to any key voltage point and keep it stable. Increase the throttle from 0% linearly in fixed steps to the maximum value allowed by the ESC. After the system output stabilizes at each throttle point, record the steady-state tension and bus current data at that throttle point through the measurement system. Complete the test of all key voltage points in sequence and summarize to obtain the full power test data of the UAV. S35. Based on the full test data of UAV power, a parameterized mathematical model describing the relationship between voltage, throttle and tension, and the relationship between voltage, throttle and current is established by using the bivariate polynomial surface fitting method or the piecewise bilinear interpolation method.

[0040] The specific process of the bivariate polynomial surface fitting method in step S35 is as follows: assuming that the tension, current, voltage, and throttle are bivariate polynomial relationships, the full test data of the UAV power is substituted into the system of overdetermined equations, and the coefficients are solved using the least squares method to obtain the tension model and the current model.

[0041]

[0042] in, Input voltage for the power system, This represents the percentage of throttle output. Let the highest polynomial order of the voltage term be . Let the highest polynomial order of the throttle term be . The polynomial order variable is the voltage term. Let be the polynomial order variable of the throttle term. The coefficients to be determined for the bivariate polynomial of the tension model are... The coefficients to be determined for the bivariate polynomial of the current model are denoted as . For the order variable of the voltage term in the current model, Let be the order variable of the throttle term in the current model. The thrust values ​​of the power system under different voltages and throttle positions. These represent the powertrain bus current values ​​under different voltages and throttle positions.

[0043] The specific process of the piecewise bilinear interpolation method in step S35 is as follows: Based on the full test data of the UAV power, construct a two-dimensional grid of voltage and throttle, and determine the voltage, throttle and tension of the four adjacent nodes of any given voltage and throttle grid cell; First, calculate the normalized coordinates for the voltage and throttle dimensions:

[0044] Then, the estimated pulling force or current value corresponding to any given voltage and throttle is calculated using the bilinear interpolation formula:

[0045] in, For the first Voltage values ​​of each grid node, For the first Voltage values ​​of each grid node, For the first Throttle value of each grid node For the first Throttle value of each grid node For the voltage dimension, normalized coordinates, For the normalized coordinates of the throttle dimension, The interpolated value represents the tension or current at the target point. For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current.

[0046] Step S4 is as follows: S41. Select either discrete specification screening mode or continuous capacity optimization mode according to project requirements. Discrete specification screening mode is suitable for commercially available battery selection scenarios, while continuous capacity optimization mode is suitable for customized cell R&D scenarios. S42. Based on the parameterized mathematical model established in S3, multi-constraint verification and iterative optimization are performed on each candidate scheme in the candidate scheme pool. In the discrete specification screening mode, the cyclic dependency problem between battery weight and total aircraft weight is solved by weight closed-loop iteration. In the continuous capacity optimization mode, the battery capacity that satisfies all constraints is searched with the goal of minimizing the total aircraft weight. S43. For candidate schemes that have passed multi-constraint verification and iterative optimization, a comprehensive score is calculated by weighting multiple dimensions such as voltage platform, range, weight efficiency, cost, power margin, and availability. The candidate scheme with the highest comprehensive score is selected as the optimal battery selection scheme.

[0047] The multi-constraint verification and iterative optimization of the discrete specification screening mode in S42 includes the following steps: A1. Determine the empty weight of the UAV, maximum mission payload, design maximum takeoff weight, target hovering time, basic or wind-resistant emergency thrust-to-weight ratio, battery discharge efficiency, and continuous power safety factor. A2. Calculate the maximum permissible battery weight, and retain specifications where the actual battery weight is less than the maximum permissible battery weight:

[0048]

[0049] in, For the maximum permissible battery weight, This is the empty weight. For maximum mission payload, To design the maximum takeoff weight, This refers to the actual battery weight. A3. Estimate the center value of battery capacity using empirical formulas, set a battery capacity range, eliminate specifications with capacities exceeding the battery capacity range, and generate a preliminary candidate specification list within the battery capacity range:

[0050] in, This is the center value of the battery capacity. The sum of the aircraft's empty weight and maximum mission load. The empirical coefficient for hovering power density is taken as 100-150 W / kg. This is the battery's nominal voltage. The target hovering time, For battery discharge efficiency, For continuous power safety factor; Battery capacity range is , The value ranges from 0.2 to 0.6. The value ranges from 2.0 to 3.0; A4. Perform a closed-loop weight iteration for each specification in the initial candidate specification list until the weight converges. The iteration process includes initializing the current weight of the battery, calculating the total weight of the aircraft, the pull force required for wind resistance or emergency response, and verifying that the battery meets the power, energy, and power constraints under the cutoff voltage.

[0051]

[0052]

[0053] in, To calculate the total weight of the aircraft, This is the current weight of the battery. The pulling force required to hover an aircraft. For the tensile strength required to resist wind or in emergencies, It is the acceleration due to gravity. For wind resistance or emergency response, the thrust-to-weight ratio is used. The dynamic constraint is that the maximum usable pulling force of the battery at cutoff voltage and 100% throttle is greater than the pulling force required for wind resistance or emergency situations. in, This is the cutoff voltage. Maximum available tensile force; The energy constraint is that the estimated range, calculated based on the hovering current at nominal voltage, is greater than the target hovering time.

[0054] in, To estimate battery life, This refers to the battery's nominal capacity. The floating current is the nominal voltage. The target hovering time; Power constraints include continuous power constraints and peak power constraints. Continuous power constraints are: Peak power constraint is ; in, The battery's continuous discharge rate. The hovering current at the cutoff voltage. This refers to the peak discharge rate of the battery. This is the maximum current at the cutoff voltage; A5. If the verification passes, calculate the precise weight of the battery:

[0055] in, For precise battery weight, For the typical mass energy density of a battery cell, if , If the convergence threshold is set to 0.1, the current weight is updated and the process returns to step A4 for re-verification; otherwise, the specification passes verification. A6. Calculate a comprehensive score for the validated specifications, and select the specification with the highest comprehensive score as the optimal solution:

[0056]

[0057]

[0058]

[0059]

[0060]

[0061]

[0062] in, ~ The weights are set according to project requirements, and the sum is 1.

[0063] The multi-constraint verification and iterative optimization of the continuous capacity optimization mode in S42 include the following steps: B1. Using the customized nominal battery capacity as the continuous optimization variable, minimizing the total weight of the aircraft as the objective function, and setting energy, continuous power, peak power and power redundancy as the core constraints;

[0064] in, To minimize the total weight of the aircraft, To match the customized battery capacity The corresponding battery weight, this value An increasing function; Energy constraints:

[0065] Continuous power constraint:

[0066] Peak power constraint:

[0067] Dynamic redundancy constraints: ; B2. The initial search range is determined by combining the upper limit of the hard constraint based on the maximum takeoff weight, the empirical estimation range based on the hovering time, and the rapid lower limit estimation based on the peak power requirement. Hard constraint upper limit The calculation formula is:

[0068] Empirical estimation of baseline capacity The calculation formula is:

[0069] in, The empirical coefficient for hovering power density is taken as 100-150 W / kg; Set the range of empirical estimation , Take 0.3-0.5, Select 2.0-3.0; The formula for calculating the fast lower bound is:

[0070] in, This represents the highest test voltage and the maximum current at 100% throttle. This refers to the peak cell rate. Determine the initial search range: Initial lower limit capacity Pick Initial maximum capacity Pick ; B3. Using the initial upper limit capacity as the input capacity, calculate the corresponding battery weight. It then verifies whether all core constraints are satisfied. If not, it determines that there is no feasible solution and returns to step S41; if satisfied, it proceeds to the next step.

[0071] B4. Use a binary search method for iterative search within the initial search range, and set the convergence precision. ε ,make , ,when First calculate the capacity at the midpoint. Then calculate the corresponding battery weight. And verify whether the capacity satisfies all core constraints; if it does, then let If not satisfied, then Iterate through the loop until the convergence accuracy is met, and output the minimum capacity that satisfies all constraints. This is the theoretical optimal value of battery capacity. ; B5. The theoretical optimal value of battery capacity is taken into the manufacturing tolerance and rounded to the standard manufacturing step of the cell manufacturer to obtain the final customized capacity.

[0072]

[0073] in, The nominal capacity is determined by taking into account manufacturing tolerances, with manufacturing tolerances set at ±5%. The final customized capacity after rounding is represented by the standard manufacturing increment of the cell manufacturer. Based on the final customized capacity, the continuous discharge current requirement, peak discharge current requirement, and maximum allowable weight of the customized battery are calculated, and a customized cell technical specification is generated.

[0074]

[0075]

[0076] in, To meet the requirements of continuous discharge current, To meet the peak discharge current requirements, This is the maximum permissible weight.

[0077] The battery cell specification sheet should include at least: Electrical specifications: nominal capacity and tolerances, nominal voltage, series and parallel connection method, operating voltage range, continuous / peak discharge rate.

[0078] Mechanical specifications: maximum weight, size limitations, interface and terminal requirements.

[0079] Performance and reliability: energy density requirements, internal resistance range, operating temperature range, cycle life, and safety certification requirements.

[0080] Verification requirements: The type of test report required (e.g., capacity test, rate discharge curve, temperature rise test).

[0081] Step S5 is as follows: S51. The optimal battery selection scheme is used to conduct actual flight tests. The safety boundary verification flight and the typical mission profile flight are completed in sequence. The safety boundary verification flight is to conduct a full-throttle vertical climb test in a safe airspace when the battery charge is about 30%. The typical mission profile flight needs to fully execute the simulated mission process of takeoff, climb, cruise, hovering, maneuvering and landing. S52. During the entire flight test, the measured data stream of battery voltage, current and time is recorded at a high frequency. S53. Compare the measured data stream with the predicted data stream of the parameterized mathematical model, and calculate the average current correction coefficient:

[0082] in, This is the current averaging correction factor. The average bus current during the critical phase of flight testing. The average bus current for the corresponding critical stage predicted by the model; S54. Feed the average current correction coefficient back to the multi-constraint verification and iterative optimization process in step S4 to adjust the parameterized mathematical model. S55. Archive all data from this selection process, including input parameters, process data, model parameters, correction coefficients, and flight verification reports, to build a design knowledge base that can be used for rapid initiation and experience sharing in similar projects in the future.

[0083] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for selecting, optimizing, and verifying batteries for multi-rotor unmanned aerial vehicles (UAVs), characterized in that, Includes the following steps: S1. Obtain the rated operating voltage range of all devices on the drone that draw power from the main battery, and calculate the system's safe voltage window; S2. Based on the system's safe voltage window, select feasible cell platforms, calculate the number of feasible strings, and construct a pool of candidate battery solutions. S3. Determine the full range of test voltages for all candidate batteries in the candidate solution pool, obtain corresponding data on voltage, throttle, tension, and current through power testing, and establish a parameterized mathematical model. S4. Based on the parameterized mathematical model, the candidate schemes are verified and iteratively optimized through multiple constraints to determine the optimal battery selection scheme; S5. Conduct flight verification based on the optimal battery selection scheme, record the measured data and correct the model, and archive all the data from this selection to build a design knowledge base.

2. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that: Step S1 is as follows: S11. Obtain and record the rated operating voltage range of all electrical devices on the drone that draw power from the main battery cart. ,in, For device indexing, This refers to the minimum rated operating voltage of the equipment. This refers to the maximum rated operating voltage of the equipment. S12, retrieve all devices The maximum value is the system's minimum operating voltage. All equipment The minimum value is the system's highest operating voltage. The system safety voltage window is obtained as follows: Furthermore, the output voltage of the candidate battery pack falls within the system's safe voltage window throughout the entire charge-discharge cycle.

3. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that: Step S2 is as follows: S21. Based on the system's safe voltage window and combined with project requirements, select feasible cell platforms and record their key parameters; S22. For each selected cell platform, calculate its feasible series number range and the single-cell operating voltage range. The lowest number of series in the battery cells. , the highest number of strings The number of feasible strings is to satisfy positive integers, in, For the index of the battery cell platform, For the first The minimum operating voltage of a single cell in this type of battery is... For the first The maximum operating voltage of a single cell in this type of battery is It is a rounding function. This is the floor function. For the first The minimum feasible number of series connections for this type of battery cell. For the first The maximum feasible number of series connections for this type of battery cell. This represents the actual number of battery cells connected in series. S23. Within the feasible number of strings for each cell platform, the highest number of strings is selected as the matching number of strings for that cell platform. Commercially available discrete capacity specifications that match each cell platform and the matching string number combination are screened and relevant information is recorded to form a structured candidate solution pool.

4. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that, Step S3 is as follows: S31. Traverse all candidate batteries in the candidate solution pool and extract the lowest cutoff voltage value among all batteries, denoted as... The highest fully charged voltage value is denoted as Then the test voltage over the entire range is ; S32. Select no fewer than 5 key voltage points within the entire test voltage range, and the key voltage points must include... and The two endpoints and the remaining intermediate points are selected according to the principle of uniform distribution or distribution near typical voltage values; S33. Construct a dedicated power test platform that includes a programmable DC power supply, the power unit under test, a measurement system, and a control and recording system; S34. Adjust the output voltage of the programmable DC power supply to any key voltage point and keep it stable. Increase the throttle from 0% linearly in fixed steps to the maximum value allowed by the ESC. After the system output stabilizes at each throttle point, record the steady-state tension and bus current data at that throttle point through the measurement system. Complete the test of all key voltage points in sequence and summarize to obtain the full power test data of the UAV. S35. Based on the full power test data of the UAV, a parameterized mathematical model describing the relationship between voltage, throttle and tension, and the relationship between voltage, throttle and current is established by using a bivariate polynomial surface fitting method or a piecewise bilinear interpolation method.

5. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 4, characterized in that, The specific process of the bivariate polynomial surface fitting method described in step S35 is as follows: assuming that the tension, current, voltage, and throttle are bivariate polynomial relationships, the full test data of the UAV power is substituted into the system of overdetermined equations, and the coefficients are solved using the least squares method to obtain the tension model and the current model. ， ， in, Input voltage for the power system, This represents the percentage of throttle output. Let the highest polynomial order of the voltage term be . Let the highest polynomial order of the throttle term be . The polynomial order variable is the voltage term. Let be the polynomial order variable of the throttle term. The coefficients to be determined for the bivariate polynomial of the tension model are... The coefficients to be determined for the bivariate polynomial of the current model are denoted as . For the order variable of the voltage term in the current model, Let be the order variable of the throttle term in the current model. The thrust values ​​of the power system under different voltages and throttle positions. These represent the powertrain bus current values ​​under different voltages and throttle positions.

6. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 5, characterized in that, The specific process of the piecewise bilinear interpolation method in step S35 is as follows: Based on the full test data of the UAV power, construct a two-dimensional grid of voltage and throttle, and determine the voltage, throttle and tension of the four adjacent nodes of any given voltage and throttle grid cell; First, calculate the normalized coordinates for the voltage and throttle dimensions: ， Then, the estimated pulling force or current value corresponding to any given voltage and throttle is calculated using the bilinear interpolation formula: ， in, For the first Voltage values ​​of each grid node, For the first Voltage values ​​of each grid node, For the first Throttle value of each grid node For the first Throttle value of each grid node For the voltage dimension, normalized coordinates, For the normalized coordinates of the throttle dimension, The interpolated value represents the tension or current at the target point. For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current, For grid nodes The corresponding measured values ​​of tension or current.

7. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that, Step S4 is as follows: S41. Select either discrete specification screening mode or continuous capacity optimization mode according to project requirements. The discrete specification screening mode is suitable for commercially available battery selection scenarios, while the continuous capacity optimization mode is suitable for customized cell R&D scenarios. S42. Based on the parameterized mathematical model established in S3, perform multi-constraint verification and iterative optimization on each candidate scheme in the candidate scheme pool. In the discrete specification screening mode, solve the cyclic dependency problem between battery weight and total aircraft weight through weight closed-loop iteration. In the continuous capacity optimization mode, search for battery capacity that satisfies all constraints with the goal of minimizing total aircraft weight. S43. For candidate schemes that have passed multi-constraint verification and iterative optimization, a comprehensive score is calculated by weighting multiple dimensions such as voltage platform, range, weight efficiency, cost, power margin, and availability. The candidate scheme with the highest comprehensive score is selected as the optimal battery selection scheme.

8. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that, The multi-constraint verification and iterative optimization of the discrete specification screening mode described in S42 include the following: A1. Determine the design parameters of the UAV; A2. Calculate the maximum permissible battery weight, and retain specifications where the actual battery weight is less than the maximum permissible battery weight; A3. Estimate the center value of battery capacity using empirical formulas, set the battery capacity range, and generate a preliminary candidate specification list within the battery capacity range: ， in, This is the center value of the battery capacity. The sum of the aircraft's empty weight and maximum mission load. The empirical coefficient for hovering power density is taken as 100-150 W / kg. This is the battery's nominal voltage. The target hovering time, For battery discharge efficiency, For continuous power safety factor; A4. Perform closed-loop iteration of weight for each specification in the initial candidate specification list until the weight converges, and verify that the battery meets the power, energy and power constraints under the cutoff voltage. If the verification is successful, calculate the accurate weight of the battery and converge. A5. Calculate a comprehensive score for the verified specifications and select the specification with the highest comprehensive score as the optimal solution.

9. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that, The multi-constraint verification and iterative optimization of the continuous capacity optimization mode described in S42 include, in sequence: B1. Using the customized nominal battery capacity as the continuous optimization variable, minimizing the total weight of the aircraft as the objective function, and setting energy, continuous power, peak power and power redundancy as the core constraints; B2. The initial search range is determined by combining the upper limit of the hard constraint based on the maximum takeoff weight, the empirical estimation range based on the hovering time, and the rapid lower limit estimation based on the peak power requirement. B3. Use the bisection method to iteratively search for the minimum capacity that satisfies the constraints within the initial search range; B4. The theoretical optimal battery capacity is factored into manufacturing tolerances and rounded to the standard manufacturing steps of the cell manufacturer to obtain the final customized capacity, and a customized cell technical specification is generated. ， ， in, The nominal capacity is determined by taking into account manufacturing tolerances, with manufacturing tolerances set at ±5%. The final customized capacity after rounding is represented by the standard manufacturing increment from the cell manufacturer.

10. The method for battery selection optimization and verification for multi-rotor UAVs as described in claim 1, characterized in that, Step S5 is as follows: S51. The optimal battery selection scheme is used to conduct actual flight tests. The safety boundary verification flight and the typical mission profile flight are completed in sequence. The safety boundary verification flight is to conduct a full-throttle vertical climb test in a safe airspace when the battery charge is about 30%. The typical mission profile flight needs to fully execute the simulated mission process of takeoff, climb, cruise, hovering, maneuvering and landing. S52. During the entire flight test, the measured data stream of battery voltage, current and time is recorded at a high frequency. S53. Compare the measured data stream with the predicted data stream of the parameterized mathematical model, and calculate the average current correction coefficient: ， in, This is the current averaging correction factor. The average bus current during the critical phase of flight testing. The average bus current for the corresponding critical stage predicted by the model; S54. Feed the average current correction coefficient back to the multi-constraint verification and iterative optimization process in step S4 to adjust the parameterized mathematical model; S55. Archive all data from this selection process, including input parameters, process data, model parameters, correction coefficients, and flight verification reports, to build a design knowledge base.

Images