Search

CN-121990207-A - Load balancing control method of unmanned aerial vehicle battery BMS system

CN121990207ACN 121990207 ACN121990207 ACN 121990207ACN-121990207-A

Abstract

The invention discloses a load balancing control method of an unmanned aerial vehicle battery BMS system, which relates to the technical field of computers and comprises the steps of collecting multisource flight states and battery parameters in real time and filtering; the method comprises the steps of integrating an open-circuit voltage method and an ampere-hour integrating method, dynamically estimating SOC of each battery cell, calculating consistency deviation, identifying a flight stage based on a lightweight decision tree, dynamically adjusting an equalization trigger threshold, carrying out single-channel equalization on SOC extremum battery cells through a bidirectional flyback topology when conditions are met, analyzing and solving optimal equalization current, introducing closed-loop energy efficiency feedback, and dynamically adjusting a driving duty ratio according to a copper loss model and an iron loss model to inhibit loss. The invention realizes quick response, high-efficiency low-noise equalization and whole machine endurance promotion under the high-dynamic working condition, the SOC consistency convergence time is obviously shortened, and the equalization energy consumption is lower than the preset proportion upper limit of the total discharge energy.

Inventors

  • LU JIANAN
  • LU HAITAO

Assignees

  • 南京跃飞智能科技有限公司

Dates

Publication Date
20260508
Application Date
20260306

Claims (9)

  1. 1. The load balancing control method of the unmanned aerial vehicle battery BMS system is characterized by comprising the following steps of: Step S1, acquiring multi-source flight state and battery parameter data, acquiring triaxial acceleration, attitude angle, motor load current, terminal voltage, temperature and current of each single battery cell of the unmanned aerial vehicle through an onboard sensor, and performing filtering processing on the original data based on a sliding time window to generate a denoised state vector; S2, dynamically calculating the deviation of the SOC and the consistency, adopting a dual-mode SOC estimation model integrating an open-circuit voltage method and an ampere-hour integration method, correcting the SOC of each battery cell by combining a temperature compensation factor, and calculating the difference between the maximum SOC and the minimum SOC as the consistency deviation; Step S3, a flight stage identification and balance trigger threshold mapping mechanism is constructed, the current flight stage is identified to be take-off, climb, hover, rapid descent or return by a lightweight decision tree algorithm, and a balance starting threshold is dynamically adjusted according to a preset stage-threshold mapping table, wherein the take-off and rapid descent stage threshold is set to be a first preset proportion, the hover stage is set to be a second preset proportion, and the return stage is set to be a third preset proportion; Step S4, light-weight equalization path planning and current optimization are executed, when equalization triggering conditions are met, a plurality of cells with highest and lowest SOC are selected to form an equalization pair, a bidirectional flyback topological structure is adopted, the optimal equalization current is solved through analysis, so that energy transfer efficiency is maximized, and meanwhile, the switching frequency is limited to be not more than a preset frequency upper limit so as to reduce electromagnetic interference; and S5, implementing closed-loop energy efficiency feedback and loss suppression, monitoring the input power and the output power of the equalization circuit in real time in the equalization process, calculating the instantaneous conversion efficiency, automatically reducing the equalization current amplitude if the efficiency is lower than a preset efficiency threshold, dynamically adjusting the driving duty ratio according to a copper loss and iron loss model of the magnetic element, and ensuring that the overall equalization energy consumption is lower than the preset proportion upper limit of the total discharge energy.
  2. 2. The load balancing control method of the unmanned aerial vehicle battery BMS system according to claim 1, wherein the sliding time window is realized by adopting a circular queue structure, the filtering algorithm is a first-order low-pass digital filter, the cut-off frequency is set to be 30Hz, the voltage abrupt change characteristic is reserved, and high-frequency noise is suppressed.
  3. 3. The load balancing control method of the battery system BMS of the unmanned aerial vehicle according to claim 1, wherein the dual-mode SOC estimation model is switched to an open-circuit voltage method when the battery cell standing time is greater than or equal to 30 seconds, a recursive ampere-hour integration method with a forgetting factor is adopted in a dynamic discharge stage, the forgetting factor takes a value of 0.995, the temperature compensation factor is obtained through a two-dimensional table lookup method, and the corresponding SOC correction amount is 0.8% every 10 ℃ interval of temperature change.
  4. 4. The method for controlling the load balancing of the unmanned aerial vehicle battery system BMS according to claim 1, wherein the depth of the lightweight decision tree algorithm is defined as 3 layers, the input characteristics comprise an acceleration module value, a pitch angle change rate and a motor current standard deviation, the reasoning process is completed through floating point comparison and array indexing, the time consumption is less than 80 microseconds, and the method can run on an embedded microcontroller with a main frequency of 120MHz in real time.
  5. 5. The load balancing control method of the unmanned aerial vehicle battery BMS system according to claim 1, wherein the number of the balancing pairs is not more than 25% of the total number of the battery cells, only one group of the balancing pairs is activated at a time of balancing operation, and the remaining balancing channels are kept in a high-impedance state for avoiding control collision and magnetic coupling interference caused by multi-channel parallelism.
  6. 6. The load balancing control method of the unmanned aerial vehicle battery BMS system according to claim 1, wherein the optimal balancing current is calculated based on an equivalent resistance of a balancing circuit and an on resistance of a switching device, the equivalent resistance of the balancing circuit and the on resistance of the switching device are obtained through off-line calibration, and the maximum allowable balancing current of a battery cell is not exceeded through limiting treatment.
  7. 7. The method for controlling the load balancing of the battery-operated system BMS of the unmanned aerial vehicle according to claim 1, wherein the balancing circuit adopts an integrated PCB layout, the volume of the magnetic element is less than 150 cubic millimeters, the switching device is a gallium nitride field effect transistor, the on-resistance is less than 10 milliohms, and the power consumption of the whole balancing module is less than 5 milliwatts in the standby state.
  8. 8. The load balancing control method of the unmanned aerial vehicle battery system BMS system according to claim 1, wherein after each flight mission is finished, a time stamp, a duration, a transfer electric quantity and an SOC convergence precision of the balancing event are uploaded to a ground station for constructing a battery health state evaluation database to support subsequent flight strategy optimization.
  9. 9. The load balancing control method of the unmanned aerial vehicle battery system BMS according to claim 8, wherein the SOC convergence accuracy is defined as the standard deviation of the SOC of each battery cell within 60 seconds after balancing, the SOC convergence accuracy is required to be smaller than 0.8%, and if the convergence accuracy exceeds the standard in three continuous tasks, the BMS self-checking flow is triggered and abnormality is reported.

Description

Load balancing control method of unmanned aerial vehicle battery BMS system Technical Field The invention relates to the technical field of computers, in particular to a load balancing control method of an unmanned aerial vehicle battery BMS system. Background With the rapid development of unmanned aerial vehicle technology, the unmanned aerial vehicle has put higher demands on the energy density, safety and cruising ability of a power battery system. The battery management system is used as a core component for guaranteeing safe and efficient operation of the unmanned aerial vehicle battery pack, and the load balancing control capability of the battery management system directly influences the service life and the flight reliability of the battery. Particularly, in a high-dynamic flight task, the battery needs to frequently bear complex working conditions such as high-rate charge and discharge, severe temperature fluctuation, voltage abrupt change and the like, and the battery forms a serious challenge for the real-time response capability, the energy efficiency management precision and the system light weight level of the BMS. The load balancing control of the unmanned aerial vehicle battery BMS system focuses on consistency maintenance of charge states of multiple battery cells, and aims to regulate energy distribution of each single battery cell in an active or passive mode and inhibit capacity attenuation and performance imbalance caused by manufacturing tolerance, aging difference or uneven thermal environment. The ideal equalization strategy realizes millisecond response and high-precision SOC synchronization on the premise of minimizing energy loss so as to support stable power supply requirements of the unmanned aerial vehicle in typical flight actions such as climbing, hovering, rapid descent and the like. However, the existing BMS balance control method has significant defects when dealing with special working conditions of the unmanned aerial vehicle. Firstly, the mainstream active equalization architecture relies on centralized computation and multi-module cooperation, and introduces a state evaluation index, a path optimization algorithm and a self-adaptive current regulation mechanism, which perform well in large energy storage scenes such as electric automobiles, but have high computation complexity, large occupation of hardware resources and are difficult to be deployed on a low-power consumption embedded platform of an onboard BMS. Secondly, the existing strategy is not fully coupled with unmanned aerial vehicle flight state parameters, so that an equilibrium decision and an actual battery are dynamically disjointed, response is lagged when voltage dip or temperature gradient is suddenly changed, and SOC deviation expansion between the battery cells cannot be restrained in time. Furthermore, the conventional balanced current control lacks of fine modeling of the conduction loss of the switching device and the efficiency of the magnetic element, so that unnecessary energy waste is caused, and the limited endurance time is further compressed. In addition, most schemes employ fixed threshold triggering equalization, neglecting differential tolerance to battery consistency for different flight phases, resulting in excessive equalization during non-critical periods, but failure due to delayed start during high risk phases. Disclosure of Invention The invention aims to provide a load balancing control method of an unmanned aerial vehicle battery BMS system, which aims to solve the problems of response lag and low energy efficiency of a traditional fixed threshold balancing strategy in a high-dynamic flight scene. In order to solve the technical problems, the invention provides the following technical scheme: the load balancing control method of the unmanned aerial vehicle battery BMS system comprises the following steps: Step S1, acquiring multisource flight state and battery parameter data in real time, namely synchronously acquiring triaxial acceleration, attitude angle, motor load current, terminal voltage, temperature and current of each single battery cell of the unmanned aerial vehicle through an airborne sensor, and carrying out filtering processing on original data based on a sliding time window to generate a denoised state vector; s2, dynamically calculating the SOC and consistency deviation, namely correcting the SOC of each battery cell by adopting a dual-mode SOC estimation model integrating an open-circuit voltage method and an ampere-hour integration method and combining a temperature compensation factor, and calculating the difference between the maximum SOC and the minimum SOC as the consistency deviation; Step S3, a flight phase identification and balance trigger threshold mapping mechanism is constructed, wherein the current flight phase is identified as take-off, climb, hover, rapid descent or return flight by a lightweight decision tree algorithm based on the flight state par