A balance car bus voltage real-time drop detection method and device

Abstract

The application is suitable for the technical field of voltage detection, and provides a balance car bus voltage real-time drop detection method and device. The method comprises the following steps: voltage synchronous collection is performed during the operation of an electric balance car to obtain an optimized voltage sequence; transient voltage drop compensation is performed on the optimized voltage sequence to generate an actual voltage sequence after compensation; a vehicle body posture parameter is collected, a historical voltage sequence is extracted based on the vehicle body posture parameter, mean value calculation is performed, and a voltage reference value of a current working condition is obtained; point-by-point deviation calculation is performed on the actual voltage sequence according to the voltage reference value to construct a voltage offset feature sequence; and a working condition adaptive window is set based on a user riding working condition to obtain a drop detection window. The application improves the accuracy of bus voltage anomaly recognition of the electric balance car, reduces the false alarm rate, and improves the vehicle operation safety and the battery service life.

Description

Technical Field

[0001] This invention relates to the field of voltage detection technology, and in particular to a method and device for real-time voltage drop detection of a balance vehicle bus. Background Technology

[0002] As a typical low-voltage DC-driven personal short-distance transportation tool, the electric self-balancing scooter's power system usually consists of a battery pack, a motor driver, and a control unit. During actual operation, the battery bus voltage will fluctuate significantly with changes in motor load, riding conditions, and battery internal resistance. Transient voltage drops are particularly likely to occur under high-load conditions such as acceleration, climbing, or sharp turns.

[0003] Existing technologies typically employ a fixed threshold method for undervoltage protection of the bus voltage, triggering a protection mechanism when the voltage falls below a preset threshold. However, this approach fails to adequately consider the dynamic load differences under varying riding conditions, easily misinterpreting normal load transient fluctuations as abnormal voltage drops. This leads to frequent power limiting or false protection triggers, impacting the overall riding experience and control stability. Traditional detection methods often rely on a single voltage signal for judgment, lacking the ability to comprehensively analyze multi-source information such as current changes, control duty cycle, and vehicle posture. This makes it difficult to accurately distinguish between short-term voltage changes caused by load disturbances and genuine voltage drops due to a decrease in battery power supply. Furthermore, in a PWM-driven environment, strong high-frequency ripple interference exists in the bus voltage, further reducing the reliability of judgments based solely on the original voltage. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a method and system for real-time voltage drop detection of a balance vehicle bus, thereby resolving at least one of the aforementioned technical issues.

[0005] In a first aspect, embodiments of this application provide a method for real-time voltage drop detection of a self-balancing scooter bus, comprising the following steps: Step S1: During the operation of the electric self-balancing scooter, voltage synchronous acquisition is performed to obtain an optimized voltage sequence; Step S2: Perform transient voltage drop compensation on the optimized voltage sequence to generate the compensated actual voltage sequence; Step S3: Collect vehicle body attitude parameters, extract historical voltage sequences based on vehicle body attitude parameters, calculate the mean value, and obtain the voltage reference value under the current operating condition; Step S4: Calculate the point-by-point deviation of the actual voltage sequence based on the voltage reference value to construct a voltage offset feature sequence; Step S5: Set an adaptive window based on the user's riding conditions to obtain a drop detection window; perform real-time detection of the voltage offset feature sequence based on the drop detection window; when the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state. Step S6: Perform adaptive undervoltage protection control based on voltage drop conditions.

[0006] Secondly, embodiments of this application provide a real-time voltage drop detection device for a self-balancing scooter bus, comprising: The voltage acquisition unit is used to synchronously acquire voltage during the operation of the electric self-balancing vehicle to obtain an optimized voltage sequence; The voltage drop compensation unit is used to perform transient voltage drop compensation on the optimized voltage sequence and generate the compensated actual voltage sequence. The voltage reference unit is used to collect vehicle body attitude parameters, extract historical voltage sequences based on vehicle body attitude parameters, perform mean calculation, and obtain the voltage reference value under the current operating condition. The deviation calculation unit is used to perform point-by-point deviation calculation on the actual voltage sequence based on the voltage reference value, and construct a voltage offset feature sequence. The drop detection unit is used to set an adaptive window based on the user's riding conditions to obtain a drop detection window; and to detect the voltage offset feature sequence in real time based on the drop detection window. When the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state. The undervoltage protection unit is used for adaptive undervoltage protection control based on voltage drop conditions.

[0007] In one approach provided by the aforementioned method and apparatus, by synchronously acquiring the bus voltage during the operation of the electric balance scooter and filtering and optimizing the acquired data, the influence of drive switch noise, sampling jitter, and transient spike interference on the voltage signal can be effectively eliminated. This improves the stability and accuracy of the bus voltage data, providing a high-quality data foundation for voltage drop detection and reducing false detections and missed detections caused by sampling noise. By performing transient voltage drop compensation on the optimized voltage sequence, the influence of normal voltage drop caused by battery internal resistance under conditions of high motor current output, rapid acceleration, and hill climbing can be eliminated. This makes the compensated actual voltage closer to the battery's actual power supply capacity, avoiding misjudging normal load fluctuations as voltage anomalies and improving the accuracy and reliability of bus voltage drop identification. By acquiring vehicle posture parameters to identify the user's riding condition and establishing a dynamic voltage reference value based on historical voltage sequences under the same conditions, the influence of different riding states on the bus voltage can be fully considered. This enables adaptive adjustment of the voltage reference reference under different conditions, overcoming the problem that fixed threshold detection methods are difficult to adapt to complex riding environments and improving the consistency of voltage detection under different operating scenarios.

[0008] By calculating the point-by-point deviation of the actual voltage sequence using a voltage reference value, a voltage offset feature sequence is constructed. This converts absolute voltage changes into relative offset changes, highlighting abnormal voltage change characteristics, enhancing the identifiability of minute voltage drops, and facilitating accurate characterization of bus voltage trends, thus improving the sensitivity of abnormal state identification. By dynamically setting the drop detection window based on the user's riding conditions and continuously detecting the voltage offset feature sequence within the corresponding window, the detection strategy can be adaptively adjusted for voltage fluctuation characteristics under different operating conditions. This effectively suppresses false alarms caused by transient disturbances and short-term fluctuations, improving the stability and accuracy of bus voltage drop state determination. By implementing adaptive undervoltage protection control after detecting a voltage drop, the vehicle's output capacity can be dynamically adjusted according to the degree of voltage anomaly, avoiding problems such as battery over-discharge, motor stalling, or abnormal controller shutdown. This improves the power supply safety and riding stability of the electric balance scooter during operation, while also extending the battery system's lifespan. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of the system structure in one embodiment of the present invention; Figure 2 This is a flowchart illustrating a method for real-time voltage drop detection of a balance vehicle bus in one embodiment of the present invention. Figure 3 yes Figure 2 A schematic diagram of the implementation process of step S1; Figure 4 yes Figure 2 A schematic diagram of the implementation process of step S2; Figure 5 This is a schematic diagram of a real-time voltage drop detection device for a balance vehicle bus in one embodiment of the present invention. Detailed Implementation 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, not all, of the embodiments of the present invention. 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.

[0011] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. It should also be understood that, as used in this specification and the appended claims, the term "and / or" refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0012] Furthermore, in the description of this invention and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0013] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of the invention include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0014] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0015] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0016] To address the problems mentioned above in the background art, this application provides a method and apparatus. The method provided by this invention can be applied to, for example... Figure 1 The system shown includes a client and a server.

[0017] In one embodiment, such as Figure 2 As shown, a method for real-time voltage drop detection of the bus of a self-balancing scooter is provided, which is then applied to... Figure 1 Taking the system in the example, the following steps are included: Step S1: During the operation of the electric self-balancing scooter, voltage synchronous acquisition is performed to obtain an optimized voltage sequence; Step S2: Perform transient voltage drop compensation on the optimized voltage sequence to generate the compensated actual voltage sequence; Step S3: Collect vehicle body attitude parameters, extract historical voltage sequences based on vehicle body attitude parameters, calculate the mean value, and obtain the voltage reference value under the current operating condition; Step S4: Calculate the point-by-point deviation of the actual voltage sequence based on the voltage reference value to construct a voltage offset feature sequence; Step S5: Set an adaptive window based on the user's riding conditions to obtain a drop detection window; perform real-time detection of the voltage offset feature sequence based on the drop detection window; when the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state. Step S6: Perform adaptive undervoltage protection control based on voltage drop conditions.

[0018] In this embodiment, during the operation of the electric balance vehicle, the bus voltage is synchronously acquired in real time by the on-board motor controller to construct a basic voltage data sequence for subsequent drop detection analysis. The bus voltage is proportionally scaled by a voltage divider resistor network and then input to the signal conditioning circuit to eliminate common-mode interference and sampling overshoot on the high-voltage side. The controller's internal analog-to-digital conversion unit then performs periodic sampling at a sampling frequency of 20kHz to 50kHz, thus covering the high-frequency ripple characteristics caused by the 10kHz to 30kHz driving PWM switching frequency. During sampling, a synchronous triggering mechanism based on the PWM cycle is introduced to ensure that the sampling points fall within the stable conduction range of the switching transistor, reducing the impact of switching transients on voltage measurement. The system uses a 1μs-level system clock to uniformly timestamp all sampled data, ensuring that the voltage sampling data, motor current, and control duty cycle signals are on the same time base. To address the high-frequency switching ripple and sampling jitter issues in the original voltage sequence, a combined digital low-pass filtering and moving average processing method is adopted. The low-pass filter coefficient is set in the range of 0.85 to 0.95 to suppress high-frequency noise while preserving the slow voltage variation trend. The final output is an optimized voltage sequence with a high signal-to-noise ratio and time consistency, which can accurately reflect the dynamic changes in the bus voltage.

[0019] After obtaining the optimized voltage sequence, transient voltage drop compensation is performed on the voltage measurement results based on the dynamic characteristics of motor load changes to eliminate non-realistic voltage drops caused by load abrupt changes. First, based on the motor phase current variation characteristics obtained through synchronous sampling, the instantaneous rate of change of current and the periodic mean change are calculated. The period scale is divided based on the PWM switching cycle from 10kHz to 30kHz to reflect load changes within different electrical cycles. In electrical theory, the transient voltage drop of the bus voltage is mainly caused by the combined effect of the battery's equivalent internal resistance and the instantaneous discharge current. Therefore, an equivalent voltage drop model is constructed, approximating the transient voltage drop as the product of current and equivalent internal resistance. Based on this, within a time alignment interval of 50μs to 100μs, the theoretical voltage drop is calculated point-by-point for the optimized voltage sequence and compensated off from the original measured voltage, thus obtaining the actual voltage sequence after removing the influence of load disturbances. To avoid numerical oscillations or overcorrection during the compensation process, a weighted smoothing constraint of 0.85–0.95 is applied to the compensation result. A first-order recursive filtering method is used to continuously correct the compensation voltage, ensuring the physical continuity of the compensation result in the time domain. The final output voltage sequence can effectively distinguish between transient changes caused by load disturbances and changes in the actual battery power supply capacity.

[0020] Vehicle attitude parameters are collected to characterize the current vehicle operating state and load characteristics. These parameters include vehicle pitch angle, wheel speed change rate, and drive output torque. The pitch angle is calculated by fusing data from a three-axis accelerometer and gyroscope using an inertial measurement unit (IMU), and attitude optimization is performed using Kalman filtering to suppress dynamic drift errors. The wheel speed change rate is calculated differentially based on encoder output within a 50μs–100μs time alignment interval. The drive torque is estimated in real time using motor phase current and an electromagnetic model. Based on these attitude parameters, a multi-dimensional riding condition feature vector is constructed. The current riding condition type is identified using a combination of threshold classification and similarity matching. These conditions include static equilibrium, acceleration, deceleration, climbing, descending, and sharp turning. Subsequently, historical operating data under the same condition category is extracted from the historical database, and the historical voltage sequences are time-aligned and resampled to meet a unified time reference of 50μs–100μs.

[0021] Based on the voltage reference value and the compensated actual voltage sequence, point-by-point difference calculation is performed with a time alignment of 50μs to 100μs to obtain a real-time voltage offset sequence, which is used to characterize the degree of deviation of the bus voltage from the theoretical reference state. At each sampling point, the current actual voltage is subtracted from the corresponding reference voltage point by point to form a time-continuous offset change sequence. A sliding window statistical method is used to smooth the offset sequence, with the window length set to 10 to 20 sampling points. The offset is denoised by calculating the local mean, local variance, and trend function to construct a voltage offset curve. On this basis, two types of key dynamic features are further extracted: the first-order slope is used to characterize the speed and direction of voltage offset change, and the cumulative offset area is used to characterize the duration and energy accumulation of the voltage deviation from the reference state. By combining the above features, a voltage offset feature sequence is constructed, which simultaneously possesses transient response capability and trend characterization capability.

[0022] The drop detection window is adaptively set based on the user's riding conditions, enabling the detection mechanism to adapt to different load dynamic conditions. A longer window is used in static equilibrium and constant speed states to improve stability, while a shorter window is used in acceleration, climbing, and sharp turns to improve response speed. The window length is in units of 50μs to 100μs time alignment intervals, with an overall range corresponding to 10 to 20 sampling points. Within the drop detection window, the voltage offset characteristic sequence is monitored and analyzed in real time. First, it is determined whether the offset remains in a positive offset state to ensure that the voltage is always below the reference level; second, the first-order change slope is calculated to determine whether it maintains a non-recovery trend, used to exclude short-term fluctuation recovery; finally, the cumulative offset area is calculated to determine whether the energy for sustained drop is sufficient. When the above three conditions are continuously and simultaneously met throughout the entire detection window, the system is determined to have entered a voltage drop state, thus effectively distinguishing between transient disturbances and actual power supply capacity degradation.

[0023] After confirming the voltage drop status, a graded adaptive undervoltage protection control strategy is executed based on the drop characteristic parameters to ensure safe system operation. First, the drop risk is classified into levels based on characteristics such as drop amplitude, duration, and recovery capability. In a low-risk state, the motor load demand is reduced by limiting instantaneous output power to prevent further voltage drop. In a medium-risk state, the transient discharge capability of the battery is suppressed by lowering the upper limit of the drive output current, thereby slowing the voltage drop rate. In a high-risk state, the high-power output channel of the drive is cut off, and only the controller core logic is powered to prevent the system from entering an unrecoverable undervoltage state. After the drop event ends, the drop amplitude, duration, voltage recovery characteristics, and corresponding operating parameters are written to the local storage module, thus forming a closed-loop adaptive protection mechanism to improve the long-term reliability of the system.

[0024] In one embodiment, such as Figure 3 As shown, step S1 specifically includes the following steps: The bus voltage, motor phase current and drive PWM duty cycle are synchronously collected during the operation of the electric balance vehicle to obtain the original electrical parameter sequence; The original electrical parameter sequence is subjected to heterogeneous timestamp alignment to obtain a time-aligned sequence. The time-aligned sequence is subjected to low-pass filtering and abnormal peak removal to obtain an optimized voltage sequence.

[0025] In this embodiment, during the operation of the electric balance vehicle, the on-board motor controller synchronously acquires the bus voltage, motor phase current, and drive signal duty cycle to construct an original electrical parameter sequence. The bus voltage is processed by a voltage divider resistor network and signal conditioning circuit, and then sampled by the controller's internal analog-to-digital converter at a sampling frequency of 20kHz to 50kHz to cover the 10kHz to 30kHz drive switching frequency range. The motor phase current is detected by a Hall current sensor or sampling resistor and is synchronously triggered for sampling with the same sampling period as the bus voltage. The drive signal duty cycle is read in real-time by the controller's internal output register and time-stamped based on the system's unified clock. All acquired data are uniformly timestamped based on a microsecond-level system timer to construct an original electrical parameter sequence containing bus voltage, motor current, duty cycle, and corresponding time information. This ensures the consistency and comparability of multi-source signals in the time dimension. Since the bus voltage, motor current sampling signals, and drive duty cycle signals originate from different acquisition channels, their update cycles and response delays differ. Therefore, a unified time reference is first established based on the controller's internal 1μs-level high-precision timer. Using a uniform time alignment interval of 50μs to 100μs, all signals are mapped to the same time axis. When aligning continuously sampled signals, a linear interpolation method based on time neighborhood is used for bus voltage and motor current to reconstruct non-integer sampled values, ensuring signal continuity on the time axis. For duty cycle signals, since they are discrete updates, a hold-type resampling method is used, maintaining the current duty cycle value within adjacent update cycles. Through this processing, the originally asynchronous multi-source heterogeneous signals are converted into a strictly time-consistent time-aligned sequence, thereby eliminating the error caused by sampling phase deviation in voltage dynamic change analysis and improving the time consistency and analysis accuracy of subsequent drop detection.

[0026] After obtaining the timing-aligned sequence, the bus voltage signal is processed to resist PWM high-frequency interference and suppress abnormal spikes to obtain a stable optimized voltage sequence. First, a first-order IIR digital low-pass filter is used to smooth the voltage signal; its discrete expression is as follows: ； The filter coefficient α is set between 0.85 and 0.95 to achieve a balance between response speed and smoothness, effectively filtering out PWM switching ripple above 10kHz. Subsequently, an abnormal spike detection mechanism based on a sliding window is introduced. The local mean and standard deviation are calculated within a window of 10–20 sampling points. When a sampling point deviates from the mean by more than three times the standard deviation, it is identified as an abnormal spike and corrected using a local mean substitution method. This spike is typically caused by switching transients, electromagnetic interference, or ADC sampling glitches. After filtering and spike removal, the optimized output voltage sequence accurately reflects the slow-changing dynamic characteristics of the bus voltage.

[0027] In one embodiment, such as Figure 4 As shown, step S2 specifically includes the following steps: The instantaneous rate of change, periodic mean change, and load change gradient of the motor phase current are calculated based on the optimized voltage sequence, and load disturbance analysis is performed to generate load disturbance parameters. The equivalent internal resistance of the battery system is calculated based on the load disturbance parameters to obtain the real-time equivalent internal resistance. Transient voltage drop compensation is performed on the optimized voltage sequence based on real-time equivalent internal resistance to generate the compensated actual voltage sequence.

[0028] In this embodiment, dynamic feature extraction is performed based on the motor phase current sequence obtained through synchronous sampling. First, the instantaneous rate of change of the motor phase current is calculated, i.e., the ratio of the current difference between adjacent sampling points to the time interval within a 50μs–100μs time alignment interval, used to characterize the intensity of transient load changes. Second, the periodic mean change is calculated. Using a PWM switching cycle of 10kHz–30kHz as the base window, the average current within each cycle is calculated and differn from the mean of adjacent cycles to reflect the slow load change trend. Furthermore, the instantaneous rate of change and the periodic mean change are combined to construct the load change gradient, which characterizes the acceleration characteristics of load changes. Based on this, a sliding window statistical analysis (window length 10–20 sampling points) is performed on the load change gradient to extract its mean and variance features, thus forming the load disturbance parameter. This parameter comprehensively reflects the variation characteristics of the motor load on both transient and steady-state time scales.

[0029] After obtaining the load disturbance parameters, an equivalent internal resistance estimation model is established based on the dynamic relationship between battery voltage and current. According to the battery equivalent circuit theory, the battery is considered as an equivalent model containing an ideal voltage source and series internal resistance, where changes in internal resistance directly affect the bus voltage drop. First, the instantaneous rate of change of current and the periodic mean change in the load disturbance parameters are used to weight and fuse the current disturbance intensity to obtain an equivalent load disturbance intensity index. Then, combined with the transient voltage drop in the optimized voltage sequence, the voltage drop ratio corresponding to a unit current change is calculated based on a time alignment of 50μs to 100μs, thereby deriving the real-time equivalent internal resistance. This equivalent internal resistance is dynamically updated using a sliding time window with a window length of 10 to 20 sampling points to avoid estimation bias caused by instantaneous noise. At the same time, a smoothing update mechanism is introduced to ensure that the internal resistance change meets the continuity constraint, thus accurately reflecting the dynamic internal resistance characteristics of the battery under different SOC and temperature conditions.

[0030] After obtaining the real-time equivalent internal resistance, transient voltage drop compensation is performed on the optimized voltage sequence to eliminate the impact of load abrupt changes on the bus voltage. Based on the battery internal resistance voltage drop model, the theoretical voltage loss caused by current changes is calculated, where the voltage loss equals the product of the real-time equivalent internal resistance and the motor phase current. This voltage drop is calculated point-by-point within a uniform time alignment interval of 50μs to 100μs and compensated from the optimized voltage sequence to obtain the compensated actual voltage sequence. During the compensation process, a smoothing constraint based on a filter coefficient of 0.85 to 0.95 is introduced to perform low-pass smoothing on the compensation result to avoid over-compensation introducing high-frequency oscillations. At the same time, a sliding window (10 to 20 sampling points) is used to verify the consistency of the voltage difference before and after compensation, ensuring that the compensation process conforms to the physical trend of voltage change. After this processing, the output actual voltage sequence can effectively eliminate the influence of transient voltage drop caused by load disturbances, so that subsequent voltage drop detection only reflects the actual power supply capacity change of the battery, rather than the transient interference of the load.

[0031] In one embodiment, step S3 specifically includes the following steps: Collect vehicle body attitude parameters; the vehicle body attitude parameters include vehicle pitch angle, wheel speed change rate and drive output torque; The riding conditions are analyzed based on the vehicle posture parameters to obtain the user's riding conditions; the riding conditions include static balance, acceleration, deceleration, climbing, descending, and sharp turning.

[0032] Based on the user's riding conditions, historical voltage detection and collection are performed to obtain historical voltage sequences; The voltage reference value for the current operating condition is obtained by averaging historical voltage sequences.

[0033] In this embodiment, vehicle attitude parameters are collected collaboratively by an onboard inertial measurement unit, wheel speed sensors, and a motor controller. These parameters include vehicle pitch angle, wheel speed change rate, and drive output torque. The pitch angle is calculated using a fusion of a three-axis accelerometer and a gyroscope, and attitude calculation is performed using complementary filtering or Kalman filtering to reduce the impact of dynamic noise. The wheel speed change rate is calculated based on the difference between adjacent samples within a 50μs–100μs time alignment interval using a wheel speed encoder. The drive output torque is estimated by the motor controller based on phase current and an electromagnetic model. A riding condition feature vector is constructed based on these multi-source attitude parameters. By setting threshold ranges and state transition logic, riding conditions are identified, classifying the current riding state into static equilibrium, acceleration, deceleration, climbing, descent, and sharp turning states, thus forming a riding condition classification result strongly correlated with power load.

[0034] Based on the currently identified riding condition type, corresponding operating condition segments are retrieved from the historical operation database to construct a historical operating condition voltage acquisition set. The historical database stores multi-dimensional operating data such as bus voltage, motor current, and drive duty cycle collected under different riding conditions, and is marked with a unified timestamp. During the retrieval process, based on the consistency of operating condition type, wheel speed change rate and pitch angle similarity are introduced for auxiliary matching, selecting the historical operating condition segment with the highest similarity. After selecting the segment, its bus voltage data is extracted and resampled at time alignment intervals of 50μs to 100μs to form a historical voltage sequence under a unified time axis.

[0035] Based on a time alignment interval of 50μs to 100μs, the historical voltage sequence is divided into multiple continuous time windows, with a window length of 10 to 20 sampling points. The voltage values ​​within each window are calculated using an arithmetic mean to obtain a local mean sequence. Subsequently, all window means are weighted and fused, with the weights determined by the similarity of operating conditions and the motor load intensity, resulting in a more representative voltage benchmark estimate. Theoretically, this benchmark value can be considered as the expected steady-state response of the bus voltage under the same operating conditions, used to eliminate the influence of individual transient disturbances. The voltage benchmark value constructed using this method can reflect the typical output characteristics of the battery system under specific riding conditions (such as acceleration, climbing, or sharp turns).

[0036] In one embodiment, step S4 specifically includes the following steps: Based on the voltage reference value, the actual voltage sequence is calculated point by point to obtain the real-time voltage offset at multiple time points; A sliding statistical analysis is performed on the real-time voltage offset to construct a voltage offset curve; The voltage offset feature sequence is constructed by calculating the first-order slope and cumulative offset area based on the voltage offset curve.

[0037] In this embodiment, after obtaining the voltage reference value and the compensated actual voltage sequence, a point-by-point difference calculation is performed on the two based on a unified time axis to construct a real-time voltage offset sequence. At a time alignment interval of 50μs to 100μs, the corresponding reference voltage value and actual voltage value are extracted for each sampling moment, and the difference between them is calculated. This difference is defined as the voltage offset, which characterizes the degree of deviation of the bus voltage from the ideal operating condition at the current moment. Theoretically, this offset can be considered as a comprehensive residual term of battery internal resistance voltage drop, load mutation, and control dynamic response. By calculating point-by-point on continuous sampling points, a voltage offset sequence varying with time can be obtained. This sequence can fully reflect the dynamic deviation characteristics of the bus voltage under different riding conditions (such as acceleration, climbing, or sharp turns).

[0038] A sliding window with a length of 10–20 sampling points is set. Based on a time alignment of 50 μs–100 μs, statistical processing is performed on the offset within the window, including calculating local mean and local smoothing values. The sliding window is then moved point by point along the time axis to form a continuously changing voltage offset curve. To suppress the influence of transient noise, a smoothing constraint mechanism is introduced during the sliding process to ensure that the curve changes satisfy the constraints of continuity and monotonicity. This voltage offset curve essentially reflects the temporal evolution of voltage deviation from the reference state. Compared to the original discrete offset sequence, it can effectively filter out high-frequency disturbances and highlight the slow-changing trend, thus making it more suitable for identifying drop trends and determining abnormal states.

[0039] After constructing the voltage offset curve, multidimensional features are extracted to form a voltage offset feature sequence. First, the first-order slope of the offset curve is calculated, i.e., the rate of change of offset between adjacent sampling points based on a time interval of 50μs to 100μs, used to characterize the speed and direction of voltage offset change. Second, the cumulative offset area is calculated, approximated by accumulating and integrating the area enclosed by the voltage offset curve and the time axis within a sliding window, thus reflecting the overall energy accumulation degree of voltage offset. Theoretically, the first-order slope corresponds to the dynamic response intensity of voltage change, while the cumulative area corresponds to the persistence and severity of voltage drop. By combining the slope and area features, a time-consistent voltage offset feature sequence can be constructed. This sequence can simultaneously describe transient changes and long-term trends, providing a highly reliable criterion basis for subsequent voltage drop classification and protection control.

[0040] In one embodiment, step S5 specifically includes the following steps: The fall detection window is obtained by setting an adaptive window based on the user's riding conditions. Voltage offset feature sequences are detected in real time based on a drop detection window. When the voltage offset feature sequences continuously meet the drop determination conditions within the drop detection window, a voltage drop state is determined. The voltage offset feature sequences maintain a positive offset state within the drop detection window; the first-order change slope continuously maintains a non-recovery trend; and the cumulative offset area exceeds a preset area threshold.

[0041] In this embodiment, the detection window length is dynamically adjusted based on the differences in the dynamic characteristics of the motor load under different operating conditions. A longer detection window is used in static equilibrium and constant speed riding states, while a shorter detection window is used in acceleration, climbing, and sharp turning states to enhance the response capability to rapid voltage changes. The window length is set based on a time alignment interval of 50μs to 100μs, and the overall window range corresponds to 10 to 20 sampling points. By introducing an operating condition weighting factor, the window length is adaptively corrected so that the detection window can match the voltage dynamic change characteristics under different load disturbance intensities, thereby constructing a drop detection window strongly correlated with riding conditions.

[0042] After obtaining the voltage drop detection window, real-time windowed detection and analysis are performed on the voltage offset feature sequence. Based on a time alignment of 50μs to 100μs, the voltage offset feature sequence is mapped to the voltage drop detection window, and its change characteristics are judged point by point. When the sequence enters the detection window, it is first determined whether its offset remains positive, i.e., the voltage is always lower than the reference value, to characterize a continuous voltage drop trend; at the same time, the first-order change slope is calculated to determine whether it maintains a non-recovery trend, i.e., the slope is always non-positive or shows a downward trend, to exclude short-term fluctuation recovery situations; in addition, the cumulative offset area is approximated by real-time integration. When it exceeds a preset area threshold, it indicates that the voltage drop has persistence and intensity characteristics. When the above three conditions are continuously and simultaneously satisfied throughout the entire voltage drop detection window, the current system is determined to have entered a voltage drop state, thereby achieving the distinction between transient disturbances and real voltage drops.

[0043] A joint constraint verification mechanism is further introduced within the voltage drop detection window to verify the stability of the voltage drop state. A sliding consistency test is performed on the voltage offset feature sequence, and a continuity analysis is conducted on each sampling point within the window to ensure that the positive offset state is uninterrupted throughout the window. Simultaneously, a trend consistency constraint is applied to the first-order change slope, requiring it to maintain a monotonous, non-recovery trend within the window to avoid misjudging local rebounds as normal states. Furthermore, a cumulative stability judgment is performed on the cumulative offset area to ensure that it shows continuous growth or remains at a high level within the window. When the three conditions satisfy the continuity and consistency constraints within the window, the final voltage drop confirmation result is output, thereby effectively distinguishing between battery power supply capacity degradation and transient load fluctuations, improving the reliability and anti-interference capability of the balance scooter bus voltage drop detection.

[0044] In one embodiment, step S6 specifically includes the following steps: Based on the voltage drop state, the voltage deviation curve is continuously tracked to extract the voltage drop amplitude, drop duration and voltage recovery rate. Voltage sag rating is performed based on voltage sag amplitude, sag duration, and voltage recovery rate to obtain a classification result; The adaptive undervoltage protection control is performed according to the graded loss. Specifically, the adaptive undervoltage protection control is as follows: under low-risk drop conditions, the instantaneous output power is limited; under medium-risk drop conditions, the upper limit of the drive output current is reduced; under high-risk drop conditions, the high-power output channel of the drive is cut off, the power supply to the core logic of the controller is maintained, and the parameters of this drop event are recorded and written to the local storage.

[0045] In this embodiment, the dynamic changes of the voltage offset curve during the drop process are continuously tracked, and three key parameters are extracted: voltage drop amplitude, drop duration, and voltage recovery rate. The voltage drop amplitude is defined as the difference between the maximum offset within the drop window and the reference voltage, used to characterize the severity of the drop. The drop duration is calculated based on the time difference between the start of the drop and the recovery determination time, reflecting the persistence of the drop event. The voltage recovery rate is calculated using the first time derivative of the voltage offset during the recovery phase, used to characterize the system's power supply recovery capability. These parameters are dynamically updated using a sliding window of 10-20 sampling points to adapt to voltage variation characteristics under different riding conditions, thus forming a complete set of drop characteristic parameters.

[0046] The voltage drop amplitude is normalized to match the system's rated voltage range. Secondly, the drop duration is time-scaled to reflect the sustained impact of voltage drops under different operating conditions. Simultaneously, the voltage recovery rate is weighted inversely to reflect the negative correlation between system recovery capability and risk. Based on a 50μs–100μs time alignment, three types of characteristic parameters are input into a weighted scoring model, and a comprehensive risk score is generated by setting different weight coefficients. The scores are categorized into three levels: low risk, medium risk, and high risk. Low risk corresponds to minor transient voltage drops, medium risk corresponds to persistent load disturbance voltage drops, and high risk corresponds to a significant decrease in battery power supply capability, thus achieving a refined classification and assessment of voltage drop events.

[0047] An adaptive undervoltage protection control strategy is implemented based on the classification results to ensure the safe operation of the electric balance vehicle system. In low-risk drop conditions, the bus load pressure is reduced by limiting instantaneous output power, keeping the motor output torque within a safe range. In medium-risk drop conditions, the transient discharge capacity of the battery is suppressed by lowering the upper limit of the drive output current, reducing the risk of further voltage drop. In high-risk drop conditions, the high-power drive output channel is cut off, retaining power only for the controller's core logic to avoid system-level instability or undervoltage protection triggering. Simultaneously, after each drop event, the drop amplitude, duration, recovery rate, and classification results are written to the local storage unit for subsequent operating condition modeling and battery health status analysis. Through this graded control mechanism, a layered response and safe closed-loop control for voltage drop risks are achieved.

[0048] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0049] In one embodiment, a real-time voltage drop detection device for the busbar of a self-balancing scooter is provided, which corresponds one-to-one with the method described in the above embodiments. For example... Figure 5 As shown, the device includes: The voltage acquisition unit is used to synchronously acquire voltage during the operation of the electric self-balancing vehicle to obtain an optimized voltage sequence; The voltage drop compensation unit is used to perform transient voltage drop compensation on the optimized voltage sequence and generate the compensated actual voltage sequence. The voltage reference unit is used to collect vehicle body attitude parameters, extract historical voltage sequences based on vehicle body attitude parameters, perform mean calculation, and obtain the voltage reference value under the current operating condition. The deviation calculation unit is used to perform point-by-point deviation calculation on the actual voltage sequence based on the voltage reference value, and construct a voltage offset feature sequence. The drop detection unit is used to set an adaptive window based on the user's riding conditions to obtain a drop detection window; and to detect the voltage offset feature sequence in real time based on the drop detection window. When the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state. The undervoltage protection unit is used for adaptive undervoltage protection control based on voltage drop conditions.

[0050] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0051] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0052] The aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), or field-programmable gate arrays (FPGAs). Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0053] The memory can be an internal storage unit of the computer device, such as a hard drive or RAM. The memory can also be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory can include both internal and external storage units of the computer device.

[0054] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0055] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0056] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0057] In the embodiments provided in this application, it should be understood that the disclosed apparatus / devices and methods can be implemented in other ways. For example, the apparatus / device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0058] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0059] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for real-time voltage drop detection of a self-balancing scooter bus, characterized in that, Includes the following steps: Step S1: During the operation of the electric self-balancing scooter, voltage synchronous acquisition is performed to obtain an optimized voltage sequence; Step S2: Perform transient voltage drop compensation on the optimized voltage sequence to generate the compensated actual voltage sequence; Step S3: Collect vehicle body attitude parameters, extract historical voltage sequences based on vehicle body attitude parameters, calculate the mean value, and obtain the voltage reference value under the current operating condition; Step S4: Calculate the point-by-point deviation of the actual voltage sequence based on the voltage reference value to construct a voltage offset feature sequence; Step S5: Set an adaptive window based on the user's riding conditions to obtain a drop detection window; perform real-time detection of the voltage offset feature sequence based on the drop detection window; when the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state. Step S6: Perform adaptive undervoltage protection control based on voltage drop conditions.

2. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 1, characterized in that, The specific steps of step S1 are as follows: The bus voltage, motor phase current and drive PWM duty cycle are synchronously collected during the operation of the electric balance vehicle to obtain the original electrical parameter sequence; The original electrical parameter sequence is subjected to heterogeneous timestamp alignment to obtain a time-aligned sequence. The time-aligned sequence is subjected to low-pass filtering and abnormal peak removal to obtain an optimized voltage sequence.

3. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 2, characterized in that, The specific steps of step S2 are as follows: The instantaneous rate of change, periodic mean change, and load change gradient of the motor phase current are calculated based on the optimized voltage sequence, and load disturbance analysis is performed to generate load disturbance parameters. The equivalent internal resistance of the battery system is calculated based on the load disturbance parameters to obtain the real-time equivalent internal resistance. Transient voltage drop compensation is performed on the optimized voltage sequence based on real-time equivalent internal resistance to generate the compensated actual voltage sequence.

4. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 3, characterized in that, Step S3 is as follows: Collect vehicle body attitude parameters; the vehicle body attitude parameters include vehicle pitch angle, wheel speed change rate and drive output torque; The riding conditions of the user are obtained by analyzing the riding conditions based on the vehicle posture parameters. Based on the user's riding conditions, historical voltage detection and collection are performed to obtain historical voltage sequences; The voltage reference value for the current operating condition is obtained by averaging historical voltage sequences.

5. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 4, characterized in that, The riding states include static balance, acceleration, deceleration, climbing, descending, and sharp turns.

6. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 4, characterized in that, The specific steps of step S4 are as follows: Based on the voltage reference value, the actual voltage sequence is calculated point by point to obtain the real-time voltage offset at multiple time points; A sliding statistical analysis is performed on the real-time voltage offset to construct a voltage offset curve; The voltage offset feature sequence is constructed by calculating the first-order slope and cumulative offset area based on the voltage offset curve.

7. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 6, characterized in that, The specific steps of step S5 are as follows: Based on the user's riding conditions, an adaptive window is set to obtain the fall detection window; The voltage offset feature sequence is detected in real time based on the drop detection window. When the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state.

8. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 7, characterized in that, The specific drop detection conditions are as follows: the voltage offset feature sequence maintains a positive offset state within the drop detection window; the first-order change slope continues to maintain a non-recovery trend; and the cumulative offset area exceeds a preset area threshold.

9. The method for real-time voltage drop detection of the busbar of a self-balancing scooter according to claim 6, characterized in that, The specific steps of step S6 are as follows: Based on the voltage drop state, the voltage deviation curve is continuously tracked to extract the voltage drop amplitude, drop duration and voltage recovery rate. Voltage sag rating is performed based on voltage sag amplitude, sag duration, and voltage recovery rate to obtain a classification result; Adaptive undervoltage protection control is implemented based on the graded borrowing method. The adaptive undervoltage protection control process specifically involves: limiting instantaneous output power under low-risk drop conditions, reducing the upper limit of drive output current under medium-risk drop conditions, cutting off the high-power output channel of the drive under high-risk drop conditions, maintaining power supply to the controller's core logic, and simultaneously recording the parameters of this drop event into local storage.

10. A device for real-time voltage drop detection on the busbar of a self-balancing scooter, characterized in that, The steps for implementing the real-time voltage drop detection method for the bus of a self-balancing vehicle as described in any one of claims 1 to 9 include: The voltage acquisition unit is used to synchronously acquire voltage during the operation of the electric self-balancing vehicle to obtain an optimized voltage sequence; The voltage drop compensation unit is used to perform transient voltage drop compensation on the optimized voltage sequence and generate the compensated actual voltage sequence. The voltage reference unit is used to collect vehicle body attitude parameters, extract historical voltage sequences based on vehicle body attitude parameters, perform mean calculation, and obtain the voltage reference value under the current operating condition. The deviation calculation unit is used to perform point-by-point deviation calculation on the actual voltage sequence based on the voltage reference value, and construct a voltage offset feature sequence. The drop detection unit is used to set an adaptive window based on the user's riding conditions to obtain a drop detection window; and to detect the voltage offset feature sequence in real time based on the drop detection window. When the voltage offset feature sequence continuously meets the drop judgment condition in the drop detection window, it is determined to be a voltage drop state. The undervoltage protection unit is used for adaptive undervoltage protection control based on voltage drop conditions.

Images