A method and system for regulating the speed of an electric motor for remote control of a vehicle

Abstract

The present application relates to the field of motor control adjustment, and particularly relates to a motor speed adjustment method and system for remotely driving a vehicle, the method comprising: acquiring and preprocessing vertical, lateral and longitudinal vibration signals of each wheel of the vehicle and driving speed data; extracting vibration abnormal feature sequences and average speed sequences through a sliding window method to analyze correlation and quantify road condition features; then quantifying wheel cooperative stability based on wheel vibration correlation and lateral / vertical energy ratio difference, obtaining a speed adjustment factor by fusing maximum road condition features and average stability; combining preset speed, transmission ratio and the like to calculate an actual target speed of the motor, and realizing accurate speed adjustment through PID closed-loop control. The present application constructs a speed adjustment factor by quantifying three-dimensional road condition features and vehicle dynamic stability, and realizes real-time adaptive adjustment of the speed by combining PID closed-loop control, thereby significantly improving the safety and stability of remote driving.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of motor control adjustment. In particular, it relates to a motor speed adjustment method and system for remotely driving a vehicle. BACKGROUND

[0002] Remote driving technology, as a vehicle control technology that is currently attracting much attention, enables users to remotely control vehicles outside the vehicle within a visible range, providing great convenience for scenarios such as parking and picking up vehicles. Remote driving technology breaks the traditional driving limit of having to be in the vehicle, simplifies the vehicle operation process in complex scenarios through remote command transmission and vehicle response, and gradually becomes an important technical direction for improving driving convenience.

[0003] During the operation of a remotely driven vehicle, motor speed adjustment plays a crucial role. Precise control of the speed directly determines the speed of the vehicle, and the speed is closely related to the safety and stability of driving. Constant speed control on flat roads cannot meet the needs of diverse road conditions. Only by flexibly and accurately adjusting the motor speed according to real-time road conditions and vehicle state can the vehicle maintain stable operation in different driving scenarios and avoid safety hazards caused by improper speed, which is a core link to ensure the reliability of remote driving.

[0004] In the traditional remote driving mode, the operator mainly relies on video and image feedback to judge the road conditions and sends speed adjustment instructions through the remote control. However, video and image feedback cannot accurately capture the actual degree of road bumps and cannot fully reflect the real-time stability state of the vehicle, which can easily lead to a mismatch between the set speed and the actual road conditions. At the same time, the existing motor speed control method lacks dynamic adaptation to real-time road conditions and vehicle dynamic state, for example: video or image is two-dimensional visual information, which cannot quantify three-dimensional road conditions (such as the amplitude and frequency of road bumps) and vehicle dynamic state (such as vehicle body inclination, abnormal wheel coordination, and side slip risk). Even if the video can see that the road has potholes, it cannot know the depth of the potholes and the intensity of the bumps; it can see the overall attitude of the vehicle body, but it cannot capture hidden instability signals such as vibration differences between wheels, making it difficult to achieve fine adjustment, increasing the risk of side slip and bumps in complex road conditions or turning operations, and seriously affecting the safety and stability of remote driving. SUMMARY

[0005] To solve the problem that in the traditional remote driving mode, the operator relies on video and image feedback to judge the road conditions and sends speed adjustment instructions, but this method cannot accurately capture the actual degree of road bumps and cannot fully reflect the real-time stability state of the vehicle, which can easily lead to a mismatch between the set speed and the actual road conditions, the present application provides solutions in the following aspects.

[0006] In a first aspect, a motor speed adjustment method for remotely driving a vehicle comprises: obtaining vertical, lateral and longitudinal vibration signals and driving speed data of each wheel of the vehicle, and preprocessing the vibration signals and driving speed data; based on the preprocessed vertical vibration signals of each wheel, using a sliding window method to divide data segments, extracting vibration abnormality features in each sliding window, and forming a vibration abnormality feature sequence, and an average driving speed corresponding to each sliding window and forming an average driving speed sequence, analyzing the correlation between the vibration abnormality feature sequence and the average driving speed sequence to distinguish the dominant factor of the vibration abnormality, and quantifying the road condition feature by combining the correlation result and the overall level of the vibration abnormality feature; based on the vibration abnormality feature sequence of each wheel, calculating the Pearson correlation coefficient of the vibration abnormality feature sequence between any two wheels, and quantifying the inter-wheel cooperative stability by combining the difference of the root mean square ratio of the lateral vibration signal and the vertical vibration signal of each wheel; selecting the maximum value of the road condition feature corresponding to all wheels and the average cooperative stability of all wheels for coupling operation to obtain a speed adjustment factor; calculating the actual target speed of the motor according to the preset target speed in the remote control instruction, the speed adjustment factor, the transmission ratio of the vehicle transmission system and the effective radius of the wheel; and dynamically correcting the deviation between the actual target speed and the current actual speed through a PID control algorithm to realize precise control of the driving speed of the vehicle.

[0007] Preferably, the vibration abnormality features in each sliding window are extracted, comprising:

[0008] Taking any wheel as a target wheel, presetting a sliding window length and a step size, calculating the coefficient of variation and the spectral flatness of the vertical vibration signal in each sliding window of the target wheel, normalizing the coefficient of variation of the amplitude of the vertical vibration signal, and taking 1 minus the spectral flatness as the frequency domain vibration abnormality degree; and taking the product of the normalized coefficient of variation and the frequency domain vibration abnormality degree as the vibration abnormality feature of each sliding window of the target wheel.

[0009] Preferably, the step of obtaining the road condition feature comprises:

[0010] Taking any wheel as a target wheel, calculating the average value of the driving speed data in each corresponding window of the target wheel, arranging the average driving speeds of all windows in time sequence to build an average driving speed sequence, building a vibration abnormality feature sequence from the vibration abnormality features of all sliding windows of the target wheel, calculating the Pearson correlation coefficient between the average driving speed sequence and the vibration abnormality feature sequence, taking 1 minus the value of the Pearson correlation coefficient as the road dominant factor, and taking the product of the road dominant factor and the average value of the vibration abnormality feature sequence of the target wheel as the road condition feature corresponding to the target wheel.

[0011] Preferably, the calculation method of the inter-wheel cooperative stability between any two wheels comprises:

[0012] Taking any two wheels as a wheel pair to be analyzed, a Pearson correlation coefficient of a vertical vibration abnormal characteristic sequence of the wheel pair to be analyzed is calculated, and the Pearson correlation coefficient is normalized to obtain a synchronism coefficient of the wheel pair to be analyzed;

[0013] An energy ratio of the lateral and vertical vibrations of the two wheels in the wheel pair to be analyzed is calculated, and normalized to obtain a consistency coefficient of the wheel pair vibration mode, wherein the energy ratio is obtained by calculating a root mean square ratio of the lateral and vertical vibration signals of the wheels.

[0014] The product of the synchronism coefficient and the consistency coefficient is taken as the cooperative stability between the wheel pair to be analyzed.

[0015] Preferably, the calculation method of the rotation speed adjustment factor comprises:

[0016] The road surface condition features corresponding to all the wheels are extracted, the maximum value in the road surface condition features is selected, 1 is subtracted from the maximum value to obtain a road surface condition correction coefficient, all the wheel pairs in two-by-two combination are traversed, the cooperative stabilities of all the wheel pairs are accumulated, and an average value of the overall cooperative stability of the vehicle is obtained, and the product of the road surface condition correction coefficient and the average value of the overall cooperative stability of the vehicle is taken as the rotation speed adjustment factor.

[0017] Preferably, the calculation method of the actual target rotation speed of the motor comprises:

[0018] The product of the preset target speed, the rotation speed adjustment factor and the transmission ratio of the vehicle transmission system is calculated in sequence, and after being divided by the wheel circumference, unit conversion is performed to obtain the actual target rotation speed.

[0019] Preferably, the preprocessing step comprises:

[0020] The vibration signals and the driving speed data are uniformly time-stamped, and a band-pass filtering technique is used to denoise the vibration signals.

[0021] In a second aspect, a motor rotation speed adjustment system for remotely driving a vehicle comprises a processor and a memory, and the memory stores computer program instructions which, when executed by the processor, implement the above-mentioned motor rotation speed adjustment method for remotely driving a vehicle.

[0022] The present application has the following effects:

[0023] 1、The present application extracts the time domain and frequency domain characteristics of the wheel vibration signal, quantifies the road condition characteristics combined with the vehicle speed data, and quantifies the cooperative stability based on the vibration correlation between the wheels and the lateral and vertical energy ratio, to realize accurate quantification of three-dimensional road conditions and vehicle dynamic state, solve the technical bottleneck that traditional two-dimensional visual feedback of video images cannot capture implicit information, and provide an objective and comprehensive basis for speed regulation.

[0024] 2、The present application constructs a speed adjustment factor by fusing the worst road condition and the overall vehicle stability, calculates the actual target speed combined with the preset target speed, and dynamically corrects the speed deviation through PID closed-loop control, so that the vehicle speed can adapt to road condition changes and vehicle stability fluctuations in real time, effectively avoiding the risk of increased side slip and bumping caused by mismatch between vehicle speed and actual road conditions and vehicle state in the traditional mode, and significantly improving the safety and stability of remote driving. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 is a method flowchart of steps S1-S4 in a motor speed regulation method for a remotely driven vehicle according to an embodiment of the present application.

[0026] Figure 2 is a structural block diagram of a motor speed regulation system for a remotely driven vehicle according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments.

[0028] Application scenario: The present solution is mainly applicable to short-distance commuting scenarios such as urban communities, parks, and rural roads, and focuses on the core needs of low-speed accurate control, complex road condition adaptation, and safe and stable driving of three-wheeled scooters (such as old-age scooters and light-duty service vehicles) in the process of remote driving.

[0029] Reference Figure 1 A motor speed regulation method for a remotely driven vehicle includes steps S1-S4, as follows:

[0030] S1: Obtain the vertical, lateral, and longitudinal vibration signals of each wheel of the vehicle and the driving speed data, and pre-process the vibration signals and driving speed data.

[0031] Three-axis vibration sensors are respectively fixed near each wheel of the vehicle, and the three-axis vibration sensors can synchronously collect vibration signals of the corresponding wheel in vertical, lateral and longitudinal directions; when installing, the coordinate system of each three-axis vibration sensor and the vehicle body coordinate system need to be accurately aligned to ensure that the collection direction of the vibration signal is consistent with the vehicle movement direction, and to avoid distortion of the physical meaning of the vibration signal due to the deviation of the coordinate system. At the same time, all three-axis vibration sensors are set to the same sampling frequency, and in this embodiment, the sampling frequency is set to 100 Hz for data collection, and the synchronization of the collection time sequence of each sensor is realized through a unified sampling clock, ensuring that the vibration signals corresponding to different wheels are completely aligned in the time dimension, providing a time synchronization basis for subsequent analysis of the vibration correlation between the wheels.

[0032] During the driving process of the remotely controlled vehicle, the vibration signals collected by all three-axis vibration sensors are received and obtained in real time; since the vibration signals contain low-frequency vehicle body motion interference (such as slight tilting and pitching motion of the vehicle body when driving at a constant speed) and high-frequency sensor self-noise (such as circuit thermal noise), these interference signals will mask the core vibration information that can reflect the actual road conditions, so the original vibration signal needs to be band-pass filtered.

[0033] In this embodiment, low-frequency interference with a frequency lower than 2 Hz and high-frequency noise with a frequency higher than 200 Hz are removed by band-pass filtering, and only the main vibration signals of the road surface in the frequency range of 2-200 Hz are retained. This frequency range is the core frequency interval of the road surface bumps, protrusions and other features causing wheel vibration, and can accurately represent the effect of road conditions on the wheels. The sampling frequency and the frequency retention range of the band-pass filter can be adaptively adjusted according to the actual application of the vehicle type (such as small passenger cars, three-wheeled vehicles), driving scenarios (such as urban roads, park paths) and sensor performance, combined with engineering experience values, to achieve the optimal signal collection effect.

[0034] Considering that the vehicle speed directly affects the strength of the wheel vibration signal, for example: the higher the speed on a flat road, the larger the wheel vibration amplitude; while on a rough road, the vibration signal strength is mainly dominated by the road conditions, and the influence of the vehicle speed is relatively weakened. If the vehicle speed factor is ignored, it is easy to misjudge the road conditions by analyzing the vibration signal alone. Therefore, while collecting the vibration signals of each wheel, the driving speed data of the vehicle needs to be synchronously obtained, and the vehicle speed sequence and the vibration signals of each wheel are accurately time-aligned based on a unified timestamp, ensuring that the vibration signal at each time can correspond to the vehicle speed at the same time.

[0035] The three-wheeled scooter is analyzed in this embodiment. The road bumping condition when the vehicle is running can be analyzed through the vertical vibration signal of the wheel to reflect the road condition characteristics. That is, when the three-wheeled vehicle runs on a flat road, the vertical vibration amplitude of the wheel is small, and the frequency is uniform. When running on a bumpy road (such as a tile joint, a pothole, or a gravel area), the wheel will be instantaneously impacted due to the road protrusion or depression, resulting in a significant increase in the vertical vibration amplitude and a chaotic frequency distribution. Through the characteristic analysis of the vertical vibration signal of each wheel, the strength and distribution of the road bumping can be intuitively quantified, and then the road condition characteristic index capable of guiding the speed regulation is formed.

[0036] The road condition characteristics are mainly quantitatively analyzed through the vertical vibration signal characteristics of each wheel. Specifically, the vertical vibration signal is directly derived from the contact impact of the wheel and the road. The amplitude, frequency, and other fluctuation characteristics of the vertical vibration signal are strongly correlated with the road bumping degree. However, when quantifying the road condition characteristics using the vertical vibration signal, the interference of the vehicle speed on the vibration signal must be considered. Since there is a significant dynamic correlation between the vehicle speed and the vertical vibration characteristics of the wheel, the vehicle speed will change with the road condition.

[0037] When the vehicle runs on a flat road, the contact state of the wheel and the road is stable, and the strength and frequency of the vibration signal are mainly determined by the vehicle speed. At this time, the vehicle speed and the vertical vibration characteristics have a high positive correlation, that is, the higher the vehicle speed, the greater the amplitude of the vertical vibration signal of the wheel and the more regular the frequency. When the vehicle runs on a rugged road (such as a pothole, a protrusion, or a gravel road), the impact of the road terrain on the wheel is much greater than the influence of the vehicle speed, becoming the core factor of the vibration signal. At this time, the interference of the vehicle speed change on the vibration signal is greatly weakened, and the correlation between the vehicle speed and the vertical vibration signal is significantly reduced. The fluctuation of the vibration signal is mainly determined by the undulating condition of the road itself. Therefore, the correlation between the vibration signal and the vehicle speed data needs to be considered to eliminate the influence of the vehicle speed on the vibration signal. The specific operation steps are as follows:

[0038] S2: Based on the preprocessed vertical vibration signal of each wheel, a sliding window method is used to divide the data section, extract the vibration abnormality characteristics in each sliding window, and form a vibration abnormality characteristic sequence. The average running speed corresponding to each sliding window is formed, and the average running speed sequence is formed. The correlation between the vibration abnormality characteristic sequence and the average vehicle speed sequence is analyzed to distinguish the dominant factor of the vibration abnormality. The correlation result and the overall level of the vibration abnormality characteristics are combined to quantify the road condition characteristics.

[0039] If the traditional Pearson correlation coefficient is directly used to calculate the overall correlation of the entire data sequence, the dynamic synchronization of the two in different driving stages cannot be accurately reflected. Since the vehicle may frequently switch between different road conditions during driving, the correlation between the vehicle speed and the vibration signal under different road conditions is significantly different. If the overall correlation is calculated based on the complete data sequence, the true correlation of the local section will be hidden. Therefore, a sliding window is used to segment the vibration signal sequence and the vehicle speed data sequence.

[0040] Taking any wheel of the vehicle as a target wheel, a preset sliding window length and a sliding step (for example, the sliding window length is set to 1 s and the sliding step is set to 0.1 s) are used to continuously slide along the time axis of the vertical vibration signal of the target wheel after preprocessing, to obtain a plurality of continuous and partially overlapping sliding windows, so as to ensure that the dynamic changes of the vertical vibration signal are fully captured.

[0041] For the vertical vibration signal in each sliding window, the amplitude of all vibration signal points in the sliding window is extracted, the mean and standard deviation of the amplitude of the vertical vibration signal in the sliding window are calculated, and the ratio of the standard deviation to the mean is taken as the coefficient of variation. The coefficient of variation is used to represent the dispersion degree of the amplitude of the vertical vibration signal in the window. The larger the value, the more intense the amplitude fluctuation of the vibration signal in the window, and the higher the time domain abnormality degree of the vibration signal.

[0042] The fast Fourier transform is performed on the vertical vibration signal in each sliding window to convert the time domain vibration signal into a frequency domain signal, and the frequency components and corresponding energy distribution of the vibration signal in the sliding window are obtained. The spectral flatness of the sliding window is calculated based on the frequency domain signal, and the ratio between the geometric mean of the frequency domain signal and the arithmetic mean of the frequency domain signal is taken as the spectral flatness. The spectral flatness is used to represent the uniformity of the spectral energy distribution of the vibration signal in the window. The closer the value is to 1, the more uniform the spectral energy distribution at each frequency, and the more stable the vibration. The closer the value is to 0, the more the spectral energy is concentrated in a few prominent frequencies, and the higher the frequency domain abnormality degree of the vibration signal.

[0043] The coefficient of variation of each sliding window is normalized by using a normalization function to obtain the normalized coefficient of variation. The purpose is to eliminate the dimensional conflict and numerical range difference of the coefficient of variation under different windows and different working conditions, to ensure that the normalized coefficient of variation is in the interval [0, 1], and to retain the change trend of the original coefficient of variation reflecting the time domain abnormality degree of the vibration.

[0044] ​The spectral flatness of each sliding window is inversely transformed, and 1 minus the spectral flatness is taken as the degree of frequency domain vibration anomaly of the sliding window. The transformed result is positively correlated with the degree of frequency domain vibration anomaly. The closer the degree of frequency domain vibration anomaly is to 1, the more significant the frequency domain vibration anomaly is. The closer the degree of frequency domain vibration anomaly is to 0, the more stable the frequency domain vibration is, which is consistent with the trend of the normalized coefficient of variation. The product of the normalized coefficient of variation of each sliding window and the degree of frequency domain vibration anomaly is taken as the vibration anomaly feature of the sliding window.

[0045] Specifically, the vibration anomaly characteristics satisfy the following relationship:

[0046] ;

[0047] In the formula, Indicates the first The first wheel The vibration anomaly characteristics of the sliding window Indicates the first The first wheel The coefficient of variation of the vertical vibration signal amplitude within a sliding window reflects the degree of dispersion of the amplitude within that window. The larger the coefficient of variation, the higher the degree of dispersion, and the greater the abnormality of the vibration signal within that sliding window. Indicates the first The first wheel Spectral flatness of the vertical vibration signal within a sliding window This represents the normalization function to avoid dimensional conflicts.

[0048] In other words, abnormal wheel vibration caused by road bumps is reflected not only in the degree of amplitude fluctuation but also in the regularity of frequency distribution. A single-dimensional feature cannot fully reflect the essence of the bumps; therefore, it is necessary to extract both the coefficient of variation in the time domain and the spectral flatness in the frequency domain. Spectral flatness is negatively correlated with the degree of vibration anomaly. Convert to positive correlation features to ensure that the changing trends of time domain and frequency domain features are consistent.

[0049] Based on the driving speed data that has been aligned with the completion time of the vibration signal in the preprocessing stage, for each sliding window of the target wheel, all driving speed data within the corresponding time period of the sliding window are extracted, the average driving speed of the sliding window time period is calculated, and the average driving speeds of all sliding windows are arranged in the time order of the sliding windows to construct an average driving speed sequence that perfectly matches the time dimension of the target wheel vibration signal. Similarly, by arranging them in the time order of the sliding windows, a vibration anomaly feature sequence of the target wheel is constructed.

[0050] Calculate the Pearson correlation coefficient between the vibration anomaly characteristic sequence of the target wheel and the corresponding average driving speed sequence. The Pearson correlation coefficient ranges from [value missing]. The pavement dominant factor is determined by subtracting the Pearson correlation coefficient from 1. The pavement dominant factor ranges from [value missing]. (In practical applications, since vehicle speed and abnormal vibration are often positively correlated, the values ​​are concentrated in...) ), The larger the value, the higher the degree to which the vibration anomaly is dominated by road conditions, and the smaller the influence of vehicle speed on the vibration anomaly; The smaller the value, the higher the degree to which the vibration anomaly is dominated by vehicle speed changes, and the smaller the influence of road conditions on the vibration anomaly.

[0051] The mean of the vibration anomaly feature sequence of the target wheel is calculated. The mean represents the overall level of vibration anomaly of the target wheel during the entire driving process. The larger the mean, the more severe the vibration impact experienced by the target wheel and the higher the potential bumpiness of the corresponding road section. The product between the road surface dominant factor and the mean of the vibration anomaly feature sequence is used as the road surface condition feature corresponding to the target wheel. The larger the road surface condition feature, the more rugged and bumpy the corresponding road surface is. The smaller the road surface condition feature, the smoother the corresponding road surface is and the more stable the driving is.

[0052] Specifically, the road surface characteristics satisfy the following relationship:

[0053] ;

[0054] In the formula, Indicates the first Road surface characteristics corresponding to each wheel Indicates the first The vertical vibration anomaly sequence of each wheel, Indicates the first The average speed sequence of each wheel The Pearson correlation coefficient between the two sequences, with a value range of [value missing]. When the correlation between vehicle speed and vibration anomaly characteristics is high, it indicates that the vibration anomaly is more affected by vehicle speed, and the road surface is relatively flat; conversely, if the correlation between vehicle speed and vibration anomaly characteristics is low, it indicates that the vibration anomaly is more affected by road surface conditions, and the road surface is poor.

[0055] By the synergistic fusion of time domain and frequency domain features, the vibration response caused by different types of road bumps (such as large amplitude potholes and high frequency small protrusions) is comprehensively covered, and compared with single dimension feature representation, the integrity and accuracy of vibration abnormal feature are significantly improved, providing more reliable basic data for road condition evaluation. At the same time, the dynamic sliding window method is used to divide the data section, and the vibration abnormal feature sequence of each wheel and the average speed sequence of the corresponding period are constructed respectively. By calculating the Pearson correlation coefficient of the two sequences, the correlation degree of speed and vibration abnormality in different driving stages is dynamically quantified. When the vibration abnormality is dominated by speed (such as flat road high-speed driving scene), the absolute value of the correlation coefficient tends to 1, and through subsequent coupling operation, the interference of this kind of vibration on road condition evaluation can be weakened; when the vibration abnormality is dominated by road condition (such as rugged road driving scene), the absolute value of the correlation coefficient tends to 0, ensuring that the road bump feature is accurately highlighted. This design effectively avoids the defect that traditional global correlation analysis cannot distinguish the correlation difference between local period speed and vibration signal, and strips off the interference of speed factor on road condition judgment, making the road condition evaluation result more in line with the actual driving scene, and significantly improving the objectivity and accuracy of the evaluation.

[0056] After completing the quantification of road condition features, there are still limitations in adjusting the speed based only on road condition information. Road conditions only reflect the complexity of the driving environment and cannot reflect the dynamic driving stability of the vehicle itself, which is directly related to the safety boundary of remote control driving, such as side slipping, tilting and other risks. Therefore, further combined with the correlation of vibration signals between wheels and the lateral vibration signals of each wheel, a quantitative evaluation system of vehicle driving stability is constructed, and then the road condition features and stability evaluation results are synergistically fused to form a speed regulation basis that takes into account environmental adaptability and self-safety, ensuring that the speed regulation is not only adapted to road conditions, but also avoids the risk of vehicle instability. The specific steps are as follows:

[0057] S3: Based on the vibration abnormal feature sequence of each wheel, the Pearson correlation coefficient of the vibration abnormal feature sequence between any two wheels is calculated, and the difference between the root mean square ratio of the lateral vibration signal and the vertical vibration signal of each wheel is quantified. The correlation stability between wheels; select the maximum value of the road condition features corresponding to all wheels and the average value of the correlation stability of all wheel pairs for coupling operation to obtain the speed adjustment factor.

[0058] From the perspective of the physical characteristics of vehicle driving, when a vehicle is in a stable driving state, such as driving in a straight line on a flat road and turning smoothly, the contact state and force conditions between each wheel and the road surface are consistent. This is reflected in the vibration signal as the vibration characteristics of each wheel, such as the vibration amplitude and frequency distribution, which are significantly correlated. However, when a vehicle shows signs of instability or is in an unstable state, such as sideslipping, tilting, or one wheel running over a protrusion, the contact state and force balance of each wheel are broken, and this vibration correlation will be greatly weakened or even disappear. By quantifying the strength of this correlation, the overall stability of the vehicle can be intuitively reflected.

[0059] Meanwhile, lateral vibration signals (vibrations parallel to the left-right direction of the vehicle) are directly related to the vehicle's lateral stability: when the vehicle sideslips or tilts, the body will generate significant lateral displacement and vibration, resulting in a significant increase in lateral vibration intensity; while during stable driving, the lateral vibration intensity is weaker and more stable. Therefore, by analyzing the energy ratio between the lateral and vertical vibration signals of each wheel, the vehicle's lateral dynamic response can be accurately captured, providing crucial support for assessing vehicle lateral stability and compensating for the inadequacy of stability analysis based solely on the correlation of vertical vibration signals.

[0060] First, extract the first wheel from the wheel pair to be analyzed. The vertical vibration anomaly sequence of the first wheel and the first wheel The vertical vibration anomaly characteristic sequences of each wheel were analyzed, and the Pearson correlation coefficient between two vertical vibration anomaly characteristic sequences was calculated. The range of the Pearson correlation coefficient is as follows: Finally, the Pearson correlation coefficient was normalized. The specific processing method is as follows: Map it to The range is used to obtain the synchronization coefficient of the wheel pair to be analyzed. The closer the synchronization coefficient is to 1, the better the synchronization of the vertical vibration signals of the two wheels and the stronger the coordination of wheel movement during vehicle operation; the closer it is to 0, the worse the vibration synchronization and the insufficient coordination of wheel movement.

[0061] Calculate the energy ratio of the lateral vibration signal to the vertical vibration signal of each of the two wheels in the wheel pair to be analyzed. That is, calculate the root mean square (RMS) values ​​of the lateral and vertical vibration signals of each wheel, and use the ratio of these RMS values ​​as the energy ratio. Then, normalize the RMS values ​​as follows: The consistency coefficient of the vibration modes of the wheel pair is obtained. The closer the consistency coefficient is to 1, the better the consistency of the vibration modes of the two wheels, and the more balanced the lateral forces on the vehicle; the closer it is to 0, the greater the difference in vibration modes, and the greater the risk of lateral instability of the vehicle. The product of the synchronicity coefficient and the consistency coefficient is used as the cooperative stability between the wheel pair to be analyzed.

[0062] Specifically, the cooperative stability between any two wheels satisfies the following relationship:

[0063] ;

[0064] In the formula, represents the cooperative stability between the first wheel and the second wheel, represents the vertical vibration abnormality feature sequence of the first wheel, represents the vertical vibration abnormality feature sequence of the second wheel, represents the vertical vibration abnormality feature sequence of the first wheel, represents the vertical vibration abnormality feature sequence of the second wheel, represents the energy ratio of the lateral vibration signal and the vertical vibration signal of the first wheel, represents the energy ratio of the lateral vibration signal and the vertical vibration signal of the second wheel, represents the energy ratio of the lateral vibration signal and the vertical vibration signal of the first wheel, represents the energy ratio of the lateral vibration signal and the vertical vibration signal of the second wheel, represents the energy ratio of the lateral vibration signal and the vertical vibration signal of the first wheel, represents the energy ratio of the lateral vibration signal and the vertical vibration signal of the second wheel, represents the Pearson correlation coefficient between the vertical vibration abnormality feature sequences of the two wheels, which can reflect the similarity of the vertical vibration signals of the two wheels. The more similar the vibration features are, the better the stability of the vehicle during driving is.

[0065] By calculating the Pearson correlation coefficient between the vertical vibration abnormality feature sequences of the two wheels in the wheel pair to be analyzed, the synchronization of the wheel movement can be accurately quantified, indicating that the more consistent the vertical vibration signal responses of the two wheels are in the same period, the stronger the wheel movement cooperation is; otherwise, the cooperation is weaker, which provides a core reference for the overall driving stability of the vehicle.

[0066] At the same time, by calculating the root mean square ratio of the lateral and vertical vibration signals of each wheel (i.e., the energy ratio of the lateral and vertical vibration signals) respectively, and further quantifying the normalized difference degree of the energy ratio between the two wheels, the relationship between the dynamic response strength of the vehicle in the lateral and vertical directions and the consistency of the vibration mode between the wheels can be effectively reflected. The smaller the difference degree is, the closer the lateral and vertical dynamic response ratio between the wheels is, and the more balanced the lateral force of the vehicle is; otherwise, the lateral force is uneven, and the vehicle has the risk of instability such as side slipping and tilting. The two-dimensional analysis comprehensively captures the dynamic balance state of the vehicle driving, and lays a solid foundation for the comprehensive quantification of the driving stability of the vehicle.

[0067] After quantifying the road condition characteristics corresponding to each wheel, removing vehicle speed interference, reflecting the bumpiness of each wheel's driving area, and quantifying the cooperative stability between all wheel pairs, it is necessary to synergistically integrate the road condition characteristics and the cooperative stability between wheels. Road condition characteristics reflect the complexity of the vehicle's driving environment (e.g., whether the road surface is rough), while cooperative stability reflects the vehicle's dynamic safety state (e.g., whether there is a risk of instability). Together, they determine the vehicle's safety boundaries and speed adaptation requirements. The speed adjustment factor is constructed by integrating these two types of quantitative indicators, and the specific steps are as follows:

[0068] Extract road condition features corresponding to all wheels, and select the maximum value among these features. The maximum value represents the worst road condition in the vehicle's driving path. Even if most road sections are flat, as long as there are local rough sections, speed adjustment is still necessary to ensure safety. As a road surface condition correction factor, its value range is: The closer the correction factor is to 1, the better the worst road conditions (the smoother the road surface), and the less the road surface restricts the engine speed; the closer the correction factor is to 0, the worse the worst road conditions (the more rugged the road surface), and the more the engine speed needs to be reduced to ensure driving safety.

[0069] Iterate through all pairwise wheel pairs, ensuring no duplicates or omissions of valid wheel pairs. For example, if the total number of wheels is 3, cover all wheel pairs. Then, the cooperative stability of all wheel pairs is summed, and the sum is divided by the total number of effective wheel pairs to obtain the mean of the overall cooperative stability of the vehicle, with a value ranging from [value missing]. The closer the mean is to 1, the better the coordinated motion of all wheel pairs and the higher the overall driving stability of the vehicle; the closer the mean is to 0, the more likely the vehicle has insufficient coordination of multiple wheel pairs and the higher the risk of instability.

[0070] The product of the road condition correction coefficient and the average value of the overall vehicle cooperative stability is used as the speed adjustment factor.

[0071] Specifically, the speed adjustment factor satisfies the following relationship:

[0072] ;

[0073] In the formula, This indicates the motor speed adjustment factor. This represents the maximum value among the characteristic values ​​of road conditions corresponding to the wheel. The maximum value represents the impact of the worst road conditions. The worse the road conditions, the greater the reduction in engine speed, and the smaller the adjustment factor. Indicates the first Road surface characteristics corresponding to each wheel This indicates the total number of wheels on the vehicle. Indicates the first The first wheel and the first Cooperative stability between individual wheels.

[0074] This represents the average of the cooperative stability between each pair of wheels. The worse the stability, the greater the reduction in speed, and the smaller the adjustment factor.

[0075] By synergistically integrating two types of core quantitative indicators to construct a comprehensive speed adjustment factor, precise adaptation and adjustment of the motor speed can be achieved: on the one hand, based on the road condition characteristics corresponding to each wheel, the maximum value is selected to represent the worst road condition in the vehicle's driving path, ensuring that the speed adjustment can cover the safety requirements of extreme rough road conditions; on the other hand, by calculating the average value of the cooperative stability between all effective wheel pairs, the dynamic balance and cooperative motion capability of the vehicle's overall driving is comprehensively characterized, ensuring that the speed adjustment responds promptly to the risk of vehicle instability.

[0076] S4: Based on the preset target vehicle speed, speed adjustment factor, transmission ratio of the vehicle transmission system, and effective wheel radius in the remote control command, calculate the actual target speed of the motor; dynamically correct the deviation between the actual target speed and the current actual speed through the PID control algorithm to achieve precise control of the vehicle speed.

[0077] The product of the preset target vehicle speed, the speed adjustment factor, and the transmission ratio of the vehicle transmission system is calculated sequentially, and then divided by the wheel circumference and converted to a unit to obtain the actual target speed.

[0078] Specifically, the actual target rotational speed satisfies the following relationship:

[0079] ;

[0080] In the formula, This indicates the actual target speed of the electric motor. Indicates the preset target speed. This indicates the speed adjustment factor of the electric motor. Indicates the gear ratio of the vehicle's transmission system. Indicates the effective radius of the vehicle's wheels. Indicates the speed unit conversion factor. This indicates the time unit conversion factor.

[0081] Through coefficients The unit of the preset target vehicle speed is converted from (km / h) to (m / s), and then converted to (r / min) through a coefficient of 60 to adapt to the actual application scenario of motor control.

[0082] The motor's current actual speed is collected by a speed sensor and compared with the target speed to obtain the real-time speed deviation. The proportional element responds quickly to the current deviation, the integral element eliminates long-term steady-state errors, and the derivative element predicts the deviation trend and suppresses it in advance. The control quantity is obtained by weighted summation of the three.

[0083] The control input is converted into a motor drive signal to dynamically adjust the output speed. Simultaneously, new actual speeds are continuously acquired, and the above steps are repeated to form a closed loop until the speed deviation converges to the allowable range, achieving precise and stable speed regulation. This step is well-known to those skilled in the art and will not be described in detail further.

[0084] This invention also provides an electric motor speed regulation system for remotely driven vehicles. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a method for regulating the speed of an electric motor in a remotely driven vehicle according to the first aspect of the present invention. The system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, the setup and functions of which are known in the art and will not be described further here.

[0085] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for adjusting the speed of an electric motor in a remotely driven vehicle, characterized in that, include: The vibration signals and driving speed data of each wheel of the vehicle in the vertical, lateral and longitudinal directions are acquired, and the vibration signals and driving speed data are preprocessed. Based on the preprocessed vertical vibration signals of each wheel, the data segment is divided using the sliding window method. Vibration anomaly features within each sliding window are extracted and a vibration anomaly feature sequence is constructed. The average driving speed corresponding to each sliding window is also used to form an average driving speed sequence. By analyzing the correlation between the vibration anomaly feature sequence and the average vehicle speed sequence, the dominant factors of vibration anomaly are distinguished. Combining the correlation results with the overall level of vibration anomaly features, the road condition characteristics are quantified. Based on the vibration anomaly feature sequence of each wheel, the Pearson correlation coefficient between any two wheels is calculated. The difference between the root mean square ratio of the lateral and vertical vibration signals of each wheel is combined to quantify the cooperative stability between wheels. The maximum value of the road condition feature corresponding to all wheels is selected and coupled with the mean value of the cooperative stability of all wheels to obtain the speed adjustment factor. Based on the preset target vehicle speed, speed adjustment factor, transmission ratio of the vehicle transmission system, and effective wheel radius in the remote control command, the actual target speed of the electric motor is calculated; the deviation between the actual target speed and the current actual speed is dynamically corrected through the PID control algorithm to achieve precise control of the vehicle speed. The calculation of the cooperative stability between any two wheels includes the following steps: Using any two wheels as the wheel pair to be analyzed, calculate the Pearson correlation coefficient of the vertical vibration anomaly characteristic sequence in the wheel pair to be analyzed, and normalize the Pearson correlation coefficient to obtain the synchronization coefficient of the wheel pair to be analyzed. The energy ratios of the lateral and vertical vibrations of the two wheels in the wheel pair to be analyzed are calculated separately and normalized to obtain the consistency coefficient of the vibration mode of the wheel pair. The root mean square ratio of the lateral vibration signal to the vertical vibration signal of the wheel is calculated to obtain the energy ratio. The product of the synchronicity coefficient and the consistency coefficient is taken as the cooperative stability between the wheelsets to be analyzed. The calculation method for the speed adjustment factor includes: Extract the road condition features corresponding to all wheels, select the maximum value among the road condition features, and use 1 minus the maximum value as the road condition correction coefficient. Iterate through all pairwise wheel pairs, accumulate the cooperative stability of all wheel pairs, and calculate the average to obtain the mean value of the overall vehicle cooperative stability. Use the product of the road condition correction coefficient and the mean value of the overall vehicle cooperative stability as the speed adjustment factor.

2. The method for adjusting the speed of an electric motor in a remotely driven vehicle according to claim 1, characterized in that, The extraction of vibration anomaly features within each sliding window includes: Taking any wheel as the target wheel, with a preset sliding window length and step size, calculate the coefficient of variation and spectral flatness of the vertical vibration signal in each sliding window of the target wheel. Normalize the coefficient of variation of the vertical vibration signal amplitude, and use 1 minus the spectral flatness as the degree of frequency domain vibration anomaly. Use the product of the normalized coefficient of variation and the degree of frequency domain vibration anomaly as the vibration anomaly feature of each sliding window of the target wheel.

3. The method for adjusting the speed of an electric motor in a remotely driven vehicle according to claim 1, characterized in that, The steps for obtaining the road surface condition features include: Taking any wheel as the target wheel, calculate the average speed data of each corresponding window of the target wheel, arrange the average speed of all windows in chronological order, and construct an average speed sequence; construct a vibration anomaly feature sequence from the vibration anomaly features of all sliding windows of the target wheel, calculate the Pearson correlation coefficient between the average speed sequence and the vibration anomaly feature sequence, and take the value of 1 minus the Pearson correlation coefficient as the road surface dominant factor; take the product of the road surface dominant factor and the mean of the vibration anomaly feature sequence of the target wheel as the road surface condition feature corresponding to the target wheel.

4. The method for adjusting the speed of an electric motor in a remotely driven vehicle according to claim 1, characterized in that, The calculation method for the actual target speed of the electric motor includes: The product of the preset target vehicle speed, the speed adjustment factor, and the transmission ratio of the vehicle transmission system is calculated sequentially, and then divided by the wheel circumference and converted to a unit to obtain the actual target speed.

5. A method for adjusting the speed of an electric motor in a remotely driven vehicle according to claim 1, characterized in that, The preprocessing steps include: The vibration signal and driving speed data are given a unified timestamp, and bandpass filtering technology is used to denoise the vibration signal.

6. A motor speed regulation system for a remotely driven vehicle, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the method for adjusting the speed of an electric motor for a remotely driven vehicle according to any one of claims 1-5.

Images