CN-121871575-B - Self-adaptive adjustment system for sweeping machine driving strategy in response to garbage overflow state

CN121871575BCN 121871575 BCN121871575 BCN 121871575BCN-121871575-B

Abstract

The invention discloses a self-adaptive adjustment system of a sweeping vehicle driving strategy for responding to a garbage overflow state, which relates to the technical field of vehicle motion control and comprises the steps of taking a probability error as a sliding mode surface variable, calculating a speed scaling factor and a speed variable, constructing an objective function and a constraint condition, using DSQP to carry out optimization solution on the objective function to obtain an execution speed and a steering angle instruction, issuing the execution speed and the steering angle instruction to a vehicle-mounted controller for execution through an API interface, realizing high-sensitivity identification of the garbage overflow state through fusion modeling of multi-source current and loading rate, triggering and self-adaptive noise updating through disturbance spectrum intensity, improving the estimation stability of the loading state, and realizing cooperative control of safety and efficiency through driving strategy optimization of overflow probability constraint.

Inventors

WANG JIXIAO
Shi Shenghu
HU ZHENQIU
JIA WEI
CHEN JIAQI
SHI SHENGLONG
WU YI
MA JI
CHEN QINYAO

Assignees

上海神舟精宜汽车制造有限公司
上海陆家嘴建设发展有限公司

Dates

Publication Date: 20260512
Application Date: 20260317

Claims (10)

1. A self-adaptive adjustment system for the driving strategy of a sweeping vehicle responding to a garbage overflow state is characterized by comprising, The disturbance sensing module is used for collecting vehicle-mounted controller data and vehicle-mounted controller calibration parameters, obtaining disturbance original observed quantity and fusion loading rate observed value, constructing an autocorrelation estimation sequence for the disturbance original observed quantity, and calculating a spectrum intensity index based on the autocorrelation estimation sequence; The noise adjusting module is used for calculating residual energy based on the fusion loading rate observation value, acquiring noise covariance through the vehicle-mounted controller, constructing a final noise updating increment by combining a spectrum intensity index, updating and judging through a sampling period, updating the final noise updating increment, cutting the upper bound and the lower bound to obtain the final noise covariance, combining a prediction state vector and a prediction covariance generated by using a UKF method to obtain innovation amount and innovation covariance, and calculating a loading rate mean value and a loading rate standard deviation; The probability correction module is used for calculating a dimensionless threshold distance based on the loading rate mean value and the loading rate standard deviation, constructing an intercept correction term based on the innovation quantity, obtaining the skewness and kurtosis through a bitwise regression equation, correcting the dimensionless threshold distance, and inputting a standard normal CDF to obtain a probability error; And the strategy optimization module is used for taking the probability error as a sliding mode surface variable, calculating a speed scaling factor and a speed variable, constructing an objective function and a constraint condition, carrying out optimization solution on the objective function by using DSQP to obtain an execution speed and steering angle instruction, and sending the execution speed and steering angle instruction to the vehicle-mounted controller for execution through an API interface.
2. The adaptive adjustment system for the sweeping vehicle driving strategy in response to the garbage overflow condition of claim 1, wherein the calculating the residual energy based on the fused load rate observation value and the obtaining the noise covariance through the vehicle-mounted controller comprises: calculating the difference value of the fusion loading rate observation value of the sampling period t-1 and the sampling period t to obtain an innovation residual error; setting productivity parameters based on a statistical analysis method, calculating a relaxation sampling period constant, calculating the product of the relaxation sampling period constant and sampling frequency, and rounding upwards to obtain window length period numbers; Reading the noise covariance of the sampling period t-1 of the vehicle-mounted controller, and calculating the product of the noise covariance and the scaling factor to obtain the noise covariance of the sampling period t; Calculating the ratio of the residual energy to the noise covariance of the sampling period t to obtain the square of the standard deviation residual, calculating the consistency loss and calculating the allowable update coefficient; and reading the calibrated maximum spectrum intensity of the vehicle-mounted controller, calculating the ratio of the spectrum intensity index to the maximum spectrum intensity, obtaining the normalized spectrum intensity, and calculating the membership degree.
3. The adaptive adjustment system for a sweeping vehicle travel strategy in response to a garbage overflow condition of claim 2, wherein said constructing a final noise update delta comprises: Based on the membership degree, calculating basic noise increment, reading the variable point sampling period and the relaxation window length of the vehicle-mounted controller, and generating an attenuation factor through exponential attenuation; And multiplying the attenuation factor, the allowable update coefficient and the basic noise increment to obtain the final noise update increment.
4. The adaptive adjustment system for a sweeping vehicle driving strategy in response to a garbage overflow condition of claim 3, wherein updating the final noise update increment and performing upper and lower bound clipping to obtain a final noise covariance comprises: If the sampling period t is smaller than or equal to the sum of the variable point sampling period and the length of the relaxation window, setting the intra-window mark as 1, otherwise, setting the intra-window mark as 0; if the noise is 0, noise adjustment is not performed, the noise covariance of the sampling period t is set to be equal to the noise covariance of the sampling period t-1, the noise covariance is marked as a frozen covariance, and if the noise covariance is 1, the noise update amount is adjusted to obtain a measured noise covariance; and performing upper and lower bound clipping on the frozen covariance and the measured noise covariance to obtain the final noise covariance.
5. The adaptive adjustment system for a sweeping vehicle travel strategy in response to a garbage overflow condition of claim 4, wherein said predicted state vector and predicted covariance generated using the UKF method comprises: reading a state estimation vector and covariance of a sampling period t-1 of the vehicle-mounted controller; Generating a predicted state vector and a predicted covariance of the sampling period t by using an unscented Kalman filter UKF method for the state estimation vector and the covariance of the sampling period t-1; mapping the prediction state vector to a measurement space through a measurement function to obtain a prediction measurement mean value.
6. The adaptive adjustment system for a sweeping vehicle travel strategy in response to a garbage overflow condition of claim 5, wherein said deriving innovation amounts and innovation covariances and calculating a load rate mean and a load rate standard deviation comprises: Calculating a difference value between the fusion loading rate observation value and the prediction measurement mean value to obtain an innovation amount, and calculating a sum of a prediction covariance and a final noise covariance to obtain an innovation covariance; Based on the innovation covariance and innovation amount, a loading rate mean and a loading rate standard deviation are calculated.
7. The adaptive adjustment system for the sweeping vehicle driving strategy in response to the garbage overflow condition of claim 6, wherein the obtaining of the skewness and kurtosis by the fractional regression equation, correcting the dimensionless threshold distance, and inputting the standard normal CDF, obtaining the probability error, comprises: Setting an overflow boundary threshold based on historical experimental experience, and dividing the difference value obtained by subtracting the loading rate mean value from the overflow boundary threshold by the loading rate standard deviation to obtain a dimensionless threshold distance; Performing standardized processing on the innovation quantity by using innovation covariance to obtain standardized innovation quantity, performing intercept correction on the standardized innovation quantity to obtain an intercept correction term, obtaining an empirical score through a score regression equation, constructing skewness and kurtosis, and correcting the dimensionless threshold distance to obtain a correction threshold distance; And inputting the correction threshold distance into a standard normal CDF to obtain a correction tail probability, and calculating the difference between the correction tail probability and the target overflow probability to obtain a probability error.
8. The adaptive adjustment system for the sweeping vehicle driving strategy in response to the garbage overflow condition of claim 7, wherein the optimizing and solving the objective function by DSQP to obtain the executing speed and steering angle command is issued to the vehicle-mounted controller for executing through the API interface, and the adaptive adjustment system comprises: Setting the probability error as a sliding mode surface variable, calculating a speed scaling factor to obtain a speed variable, constructing an objective function, setting constraint conditions, carrying out optimization solving on the objective function by using DSQP to obtain an execution speed and steering angle instruction, and issuing and executing the execution speed and steering angle instruction through an API interface.
9. The adaptive adjustment system for a sweeping vehicle driving strategy in response to a garbage overflow condition of claim 8, wherein said constructing an autocorrelation estimation sequence for an original observed quantity of disturbance, and calculating a spectral intensity index based on the autocorrelation estimation sequence, comprises: Setting the length of a fixed window as N, collecting disturbance original observed quantity in the fixed window to obtain a disturbance window sequence with mean value removed, calculating an autocorrelation estimation sequence, and calculating a power spectral density function value; and setting a frequency band set based on engineering calibration in the vehicle-mounted controller, and summing the power spectral density function values in the frequency band set to obtain a spectral intensity index.
10. The adaptive adjustment system for the sweeping vehicle driving strategy in response to the garbage overflow condition of claim 9, wherein the collecting the vehicle-mounted controller data and the vehicle-mounted controller calibration parameters to obtain the disturbance original observed quantity and the fusion loading rate observed value comprises the following steps: and acquiring data of the vehicle-mounted controller and calibration parameters of the vehicle-mounted controller through an API (application program interface), calculating a loading rate variable and a height channel loading rate, and carrying out weighted summation to obtain a fusion loading rate observation value.

Description

Self-adaptive adjustment system for sweeping machine driving strategy in response to garbage overflow state Technical Field The invention relates to the technical field of vehicle motion control, in particular to a self-adaptive adjustment system for a sweeping vehicle driving strategy in response to a garbage overflow state. Background The existing road sweeper realizes basic monitoring of a sweeping state and a loading state through a vehicle speed sensor, a dustbin liquid level or quality sensor, a motor current sensor and the like, completes sweeping, dust collection and garbage collection operations by combining preset operation parameters, introduces a vehicle-mounted controller and a background management platform, realizes uploading of the operation state, abnormal alarming and remote adjustment of an operation route through a communication interface, achieves certain progress in improving the operation efficiency and reducing manual intervention, and becomes an important development direction of current intelligent sanitation equipment. However, the prior art still has the defects that the prior loading state estimation method adopts a fixed noise parameter adjustment mode, the disturbance energy frequency domain characteristics in the operation process cannot be analyzed, so that excessive response or long-term deviation easily occurs after the state estimation occurs, quick and stable estimation recovery is difficult to realize after the disturbance, the prior sweeping machine driving strategy adjustment is triggered based on rules, the garbage overflow risk is not introduced into a control target in a probability mode, and the cooperative optimization control of speed and steering is difficult to realize while the stability, the safety and the operation efficiency of the vehicle are ensured. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides a self-adaptive adjustment system for the sweeping vehicle driving strategy responding to the garbage overflow state, which solves the problems that the existing loading state estimation method adopts a fixed noise parameter adjustment mode, the disturbance energy frequency domain characteristics in the operation process cannot be analyzed, so that the state estimation is easy to generate excessive response or long-term deviation after the disturbance occurs, the rapid and stable estimation recovery is difficult to realize after the disturbance, the existing sweeping vehicle driving strategy adjustment is triggered based on rules, the garbage overflow risk is not introduced into a control target in a probability mode, and the cooperative optimization control of speed and steering is difficult to realize while the stability, the safety and the operation efficiency of the vehicle are ensured. In order to solve the technical problems, the invention provides the following technical scheme: The invention provides a self-adaptive adjustment system of a sweeping machine driving strategy responding to a garbage overflow state, which comprises the following components; The disturbance sensing module is used for collecting vehicle-mounted controller data and vehicle-mounted controller calibration parameters, obtaining disturbance original observed quantity and fusion loading rate observed value, constructing an autocorrelation estimation sequence for the disturbance original observed quantity, and calculating a spectrum intensity index based on the autocorrelation estimation sequence; The noise adjusting module is used for calculating residual energy based on the fusion loading rate observation value, acquiring noise covariance through the vehicle-mounted controller, constructing a final noise updating increment by combining a spectrum intensity index, updating and judging through a sampling period, updating the final noise updating increment, cutting the upper bound and the lower bound to obtain the final noise covariance, combining a prediction state vector and a prediction covariance generated by using a UKF method to obtain innovation amount and innovation covariance, and calculating a loading rate mean value and a loading rate standard deviation; The probability correction module is used for calculating a dimensionless threshold distance based on the loading rate mean value and the loading rate standard deviation, constructing an intercept correction term based on the innovation quantity, obtaining the skewness and kurtosis through a bitwise regression equation, correcting the dimensionless threshold distance, and inputting a standard normal CDF to obtain a probability error; And the strategy optimization module is used for taking the probability error as a sliding mode surface variable, calculating a speed scaling factor and a speed variable, constructing an objective function and a constraint condition, carrying out optimization solution on the objecti