A control method for an information-based weighing system for electric sanitation and kitchen waste trucks

Abstract

This invention relates to the field of intelligent weighing technology, and more particularly to a control method for an information-based weighing system for electric sanitation and catering vehicles. The method includes: synchronously collecting weight data, posture data, vibration data, and pressure data during vehicle operation; classifying and identifying the vehicle's operating conditions; and determining whether high-precision weighing conditions are met. When met, a stable time window is locked through joint analysis of pressure change trends and accumulated vibration energy. Within this window, gravity component compensation, noise reduction, and outlier removal are performed on the weight data to obtain an accurate weight value. When not met, a simplified data processing method is used to obtain a weight estimate. The weighing results are stored and uploaded to a remote server in an orderly manner according to task priority rules. This invention solves the problems of low weighing accuracy and poor data reliability of electric sanitation and catering vehicles under complex dynamic operating conditions, and achieves high-precision acquisition, stable processing, and efficient transmission of vehicle load information.

Description

Technical Field

[0001] This invention relates to the field of intelligent weighing technology, and in particular to a control method for an information-based weighing system for electric sanitation and kitchen waste trucks. Background Technology

[0002] With the acceleration of modern urbanization, sanitation operations and food waste treatment play a crucial role in improving urban environmental quality. Electric sanitation and food waste collection vehicles, as an environmentally friendly and efficient urban waste disposal tool, have been widely used in urban waste collection and transportation. However, during the operation of these electric sanitation and food waste collection vehicles, due to the complexity and dynamism of the operating environment, such as uneven road surfaces, varying vehicle loads, and environmental vibrations, traditional weighing systems often struggle to adapt to these complex conditions, severely impacting the accuracy and stability of weighing data.

[0003] Existing weighing systems mostly rely on a single weight sensor, which presents numerous problems during dynamic operation. For example, when vehicles experience significant vibration or tilting, the weighing data often shows large errors, failing to accurately reflect the vehicle's actual load. Especially in scenarios with significant pressure variations, traditional weighing systems cannot effectively compensate for pressure fluctuations, leading to inaccurate weighing results and severely impacting weight monitoring and subsequent data processing during waste collection. Furthermore, most current weighing systems lack data interaction capabilities with external systems, resulting in low efficiency in data uploading and remote monitoring, hindering comprehensive real-time monitoring and precise scheduling of sanitation vehicles. This limits resource allocation and optimized management in sanitation operations, preventing the full realization of the role of electric sanitation and food service vehicles in improving urban sanitation efficiency.

[0004] Therefore, to address the aforementioned issues, this invention is proposed. By integrating multiple data sources, such as weight, posture, vibration, and pressure data, and employing high-precision algorithms for dynamic compensation and optimization, it achieves accurate monitoring of the load status of electric sanitation and food processing vehicles. This method overcomes the shortcomings of traditional weighing systems under complex environmental conditions, providing more stable and accurate weighing data. Furthermore, through information management, it enables efficient data uploading and remote monitoring, contributing to improved automation and intelligence in urban sanitation management. Summary of the Invention

[0005] This invention provides an information-based weighing system and control method for electric sanitation and kitchen waste trucks, which improves weighing accuracy under complex environmental conditions.

[0006] In a first aspect, the present invention provides a control method for an information-based weighing system of an electric sanitation and kitchen waste vehicle, the method comprising:

[0007] Step S1: Collect weight data, posture data, vibration data, and pressure data of the electric sanitation and kitchen waste vehicle during operation; determine the current vehicle operating condition information based on the weight data and posture data;

[0008] Step S2: Determine whether the high-precision weighing conditions are met based on the vehicle operating condition information; if the high-precision weighing conditions are met, lock the stable time window based on the determination result of whether the change trend of the pressure data has entered the stable range and the comparison result of the energy accumulation value of the vibration data with the preset threshold; within the stable time window, perform gravity component compensation, noise reduction processing and outlier removal on the weight data to obtain the accurate weight value.

[0009] Step S3: When the high-precision weighing conditions are not met, the attitude data, vibration data, and pressure data in the current time period are matched with the pre-built error model library to obtain the corresponding error type; the attitude data, vibration data, and pressure data are assigned corresponding weights according to the error type, and the corresponding weight estimate is obtained by combining the fusion estimation algorithm.

[0010] Step S4: Store the precise weight value or the weight estimate, and upload it to the remote server according to the upload order determined by the preset task priority rules.

[0011] As a preferred embodiment of the present invention, step S1, determining the current vehicle operating condition information, includes:

[0012] The ratio of the weight data to the vehicle's unloaded weight is calculated, and the load status is determined to be heavy or light based on a first comparison result of the ratio and a preset ratio threshold. The vehicle's tilt angle is calculated based on the attitude data, and the degree of tilt is determined to be tilted or stable based on a second comparison result of the tilt angle and a preset tilt threshold. The vehicle's operating condition is determined to be one of four types: heavy-load stable, heavy-load tilted, light-load stable, or light-load tilted, based on the load status and the degree of tilt. Under the vehicle operating condition type, a fast Fourier transform is performed on the vibration data to obtain the dominant vibration frequency band information. The peak value of the dominant vibration frequency band information and the peak value of the pressure data are time-series compared according to timestamps. The synchronization degree between the two within a preset sliding window is calculated and compared with a synchronization judgment threshold to determine whether there is an abnormal synchronization phenomenon between vibration and pressure. When the abnormal synchronization phenomenon between vibration and pressure exists, the current moment is marked as a potential abnormal operating condition point, and the corresponding timestamp and the vehicle operating condition information are recorded.

[0013] As a preferred embodiment of the present invention, in step S2, obtaining the accurate weight value includes:

[0014] Based on the vehicle operating condition information, it is determined whether the high-precision weighing conditions are met. These conditions include at least a stable pressure data trend and a cumulative vibration energy value below a preset threshold. When met, the rate of change between adjacent sampling points is calculated based on the pressure data, and it is determined whether the pressure data has entered a plateau period. Simultaneously, frequency domain analysis is performed on the vibration data, and the cumulative vibration energy value is calculated. The cumulative vibration energy value is compared with the preset threshold. When the pressure trend is stable and the cumulative vibration energy value is below the preset threshold, the current time period is locked as a stable time window. Within the stable time window, multiple sets of weight data are continuously collected at a preset sampling frequency. Gravity component compensation processing is performed on the multiple sets of weight data based on the corresponding attitude data. Noise reduction and outlier removal processing are performed on the compensated weight data. Statistical calculations are then performed on the processed weight data to obtain the accurate weight value.

[0015] As a preferred embodiment of the present invention, in step S2, obtaining the gravity compensation component includes:

[0016] After locking the stable time window based on the pressure data change trend and the cumulative vibration energy value, the current tilt angle of the vehicle is first calculated based on the attitude data. Then, according to the projection relationship of gravity on the sensor measurement direction, the original weight measurement value obtained from the pressure signal is corrected by the gravity component to obtain the initial compensation weight value. The initial compensation weight value is then time-aligned and normalized with the vibration signal characteristics, tilt angle characteristics, and pressure change characteristics within the corresponding time period. This is then used as a multi-dimensional input feature vector and input to a pre-trained small neural network model. The small neural network model performs nonlinear prediction on the weight compensation residual and outputs the weight compensation correction amount. The weight compensation correction amount is then superimposed on the initial compensation weight value to obtain the final weight compensation result.

[0017] As a preferred embodiment of the present invention, step S3, when the high-precision weighing conditions are not met, includes obtaining the corresponding weight estimate, including:

[0018] When high-precision weighing conditions are not met, attitude data, weight data, vibration data, and pressure data are collected in real time within the current time period. The tilt angle change rate, vibration energy level, pressure change rate, and their time synchronization characteristics are extracted to construct a dynamic working condition feature vector. This dynamic working condition feature vector is matched with a preset dynamic working condition mapping relationship model library. Based on the matching results, multidimensional interpolation is used to calculate the fluctuation range and offset trend of the weight data under the current dynamic working condition. The mapping relationship model library is established hierarchically based on historical calibration data, categorized by vibration level, tilt angle range, and load range. Dynamic correction coefficients for attitude data, vibration data, and pressure data are calculated based on the fluctuation range and offset trend of the weight data. These coefficients are then normalized and their change rate constrained to form corresponding dynamic weights. Within a preset sliding time window, the real-time acquired attitude data, vibration data, pressure data, and corresponding dynamic weights are input into the weight estimation model for fusion calculation. The output results are then time-smoothed to obtain the weight estimate under the current dynamic working condition.

[0019] As a preferred embodiment of the present invention, the construction process of the mapping relationship model library includes:

[0020] By operating the vehicle under different vibration levels, tilt angle ranges, and load ranges, and simultaneously collecting attitude data, vibration data, and pressure data, the fluctuation amplitude and offset of the weight measurement value converted from pressure relative to the actual weight value are calculated within the corresponding unstable time period. Operating condition characteristic parameters, including at least the tilt angle change rate, vibration energy level, and pressure change rate, are extracted. A one-to-one correspondence is established between these operating condition characteristic parameters and the corresponding weight fluctuation amplitude and offset, and the data is stratified and categorized according to vibration level, tilt angle range, and load range to form a multidimensional mapping data table. The multidimensional mapping data table is then interpolated and smoothed to generate a continuously queryable mapping relationship model library.

[0021] As a preferred embodiment of the present invention, step S3 further includes:

[0022] After obtaining the weight estimate, the most recent accurate weight value obtained within a stable time window is retrieved. The difference and relative error between the weight estimate and the accurate weight value are calculated to obtain a comparison result. When the difference or relative error is within a preset error range, the parameters in the fusion estimation algorithm are corrected using the accurate weight value as a calibration benchmark. The parameters include at least one of attitude correction coefficient, vibration suppression coefficient, pressure trend constraint coefficient, or fusion weight. The corresponding parameters are adjusted by scaling or bias compensation to make subsequent weight estimates converge to the accurate weight value. When the difference exceeds the preset error range, the current fusion estimation algorithm parameters are kept unchanged or the parameter update magnitude is reduced.

[0023] Secondly, the present invention provides an information-based weighing system control system for electric sanitation and kitchen waste trucks, used to implement the above-mentioned method, the system comprising:

[0024] The data acquisition unit is used to collect weight data, posture data, vibration data, and pressure data of the electric sanitation and food processing vehicle during operation; and to determine the current vehicle operating condition information based on the weight data and posture data.

[0025] The precision measurement unit is used to determine whether the high-precision weighing conditions are met based on the vehicle operating condition information. If the high-precision weighing conditions are met, the unit locks the stable time window based on the determination result of whether the change trend of the pressure data has entered the stable range and the comparison result of the energy accumulation value of the vibration data with the preset threshold. Within the stable time window, the weight data is subjected to gravity component compensation, noise reduction processing and outlier removal to obtain an accurate weight value.

[0026] A simplified weighing unit is used to obtain a weight estimate by using a simplified processing method if the high-precision weighing conditions are not met.

[0027] The data storage and uploading unit is used to store the precise weight value or the weight estimate, and upload it to the remote server after determining the upload order according to the preset task priority rules.

[0028] Thirdly, the present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the above-described method.

[0029] The beneficial effects of this invention are as follows:

[0030] This invention achieves refined identification of the current operating condition by simultaneously collecting multi-source data such as weight, attitude, vibration, and pressure during vehicle operation. This analysis comprehensively examines the vehicle's load state, tilt state, and vibration disturbance characteristics. Furthermore, the temporal synchronicity of vibration and pressure is used to identify potential abnormal conditions, providing reliable criteria for subsequent weighing mode selection. This establishes a basis for separating stable and unstable operating conditions, avoiding the blind output of high-precision results when conditions are not met. When high-precision weighing conditions are met, a stable time window is locked by jointly determining the pressure change trend and vibration energy threshold. Within this window, gravity component compensation, noise reduction, and outlier removal are implemented. A small neural network is used to nonlinearly correct residual errors, thereby obtaining a precise weight value close to the effect of static weighing. The collaborative compensation mechanism of the physical model and the data-driven model improves... This method ensures measurement accuracy under slightly disturbed environments. When high-precision conditions are not met, it does not simply abandon weighing; instead, it matches dynamic working condition feature vectors with a pre-defined mapping model library, obtains the fluctuation range and offset trend through multi-dimensional interpolation, and outputs a weight estimate using a dynamic weight fusion estimation algorithm. This enables continuous tracking of the weighing status under complex dynamic working conditions. Simultaneously, by comparing the estimated weight with the accurate weight value within a subsequent stable window, the fusion parameters are dynamically corrected, gradually converging the estimation results and preventing long-term error accumulation. Finally, through local data storage and task priority scheduling mechanisms, the method ensures orderly and secure uploading of weighing data under different network environments, prioritizing the transmission of safety alarm information and ensuring stable feedback of operational data. This method achieves high-precision, robust, and highly reliable load information management in complex operating environments of electric sanitation and catering vehicles. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart illustrating the control method of an information-based weighing system for an electric sanitation and kitchen waste vehicle, as shown in the embodiment.

[0033] Figure 2 This is a comparison chart of the gravity compensation residual correction effects in the embodiments;

[0034] Figure 3 This is a structural diagram of an information-based weighing system for an electric sanitation and kitchen waste truck, as shown in the embodiment. Detailed Implementation

[0035] This invention provides a control method for an information-based weighing system for electric sanitation and kitchen waste trucks. The terms "first," "second," "third," "fourth," etc. (if applicable) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0036] For ease of understanding, the specific process of the embodiments of the present invention will be described below, such as... Figure 1 As shown, an embodiment of the present invention provides a control method for an information-based weighing system for electric sanitation and kitchen waste trucks, comprising:

[0037] Step S1: Collect weight data, posture data, vibration data, and pressure data during the operation of the electric sanitation and food processing vehicle; determine the current vehicle operating condition information based on the weight data and posture data; specifically including:

[0038] The system calculates the ratio of the average weight data within a set time period to the vehicle's unloaded weight, and determines whether the load status is heavy or light based on a first comparison result between the ratio and a preset ratio threshold. It also calculates the pressure data variation amplitude within the set time period and compares it with a preset amplitude to obtain a second comparison result, determining whether the swaying degree is bumpy or stable based on the second comparison result. The system performs a Fast Fourier Transform on the vibration data to obtain the dominant vibration frequency band information, aligns the peak value of the dominant vibration frequency band information with the peak value of the pressure data according to timestamps, calculates the synchronization degree of the two within a preset sliding window, and compares it with a synchronization judgment threshold to determine whether there is an abnormal synchronization phenomenon between vibration and pressure. When an abnormal synchronization phenomenon between vibration and pressure exists, the current time is marked as a potential abnormal operating condition point, and the timestamps corresponding to the weight data, attitude data, vibration data, and pressure datasets corresponding to non-potential abnormal operating condition points are used as the vehicle operating condition information.

[0039] Specifically, multiple key data points are collected during the operation of electric sanitation and food processing vehicles, including weight data, posture data, vibration data, and pressure data. These data are collected in real time using different sensors. Weight data is obtained by converting pressure data acquired by a pressure sensor, posture data is measured by an angle sensor, and vibration and pressure data are collected by a vibration sensor and a pressure sensor, respectively, and are used to monitor the dynamic and weighing status of the vehicle.

[0040] Based on the collected weight and attitude data, the vehicle's operating condition is first determined. Specifically, the ratio of the weight data to the vehicle's unloaded weight is calculated and compared to a preset ratio threshold to determine whether the vehicle is heavily loaded or lightly loaded. A ratio greater than the threshold indicates a heavy load, and vice versa. Next, based on the attitude data, the vehicle's tilt angle is calculated and compared to a preset tilt threshold to determine whether the vehicle is tilted or stable. A tilt angle greater than the threshold indicates a tilted state, and vice versa. Based on this, and considering the combination of load condition and tilt degree, the vehicle's operating condition is further classified into one of four types: heavy-load stable, heavy-load tilted, light-load stable, or light-load tilted. Furthermore... The corresponding vibration and pressure data are processed using Fast Fourier Transform (FFT) to extract the dominant frequency band information of the vibration signal. The frequency domain information of the vibration data can effectively reflect the mechanical disturbance characteristics of the vehicle during operation. The peak value of the extracted dominant frequency band information of the vibration is compared with the peak value of the pressure data in time sequence, and the two data are aligned according to the timestamp to calculate the synchronization degree of vibration and pressure. In this process, a preset sliding time window is set to evaluate the synchronization degree of the dominant frequency band of vibration and the peak value of pressure, and compare it with the synchronization judgment threshold to determine whether there is an abnormal synchronization phenomenon between vibration and pressure. If the calculation results show that the peak values ​​of vibration and pressure appear in the same time window and the synchronization degree exceeds the preset synchronization judgment threshold, then an abnormal synchronization phenomenon between vibration and pressure is considered to exist.

[0041] Once an abnormal synchronization phenomenon is identified, the current moment is marked as a potential abnormal operating condition point, and the corresponding timestamp and vehicle operating condition type are recorded to ensure timely identification and response when abnormal vehicle operating conditions occur. To ensure the reliability and real-time performance of data processing, the characteristics of each data source are utilized to enable effective fusion and processing of different types of sensor data within a unified framework. Specifically, sensor data is first preprocessed in the time domain, such as removing noise through signal filtering, and then key features, such as vibration main frequency band information, are extracted through frequency domain analysis. This ensures that even under complex operating conditions such as vibration, tilt, pressure fluctuations, and other interference factors, effective information can still be accurately extracted from multi-source data for operating condition judgment and anomaly detection. Among these, weight data is used... To determine the load status, attitude data is used to determine the degree of tilt, vibration data is used to analyze the disturbance characteristics of the vehicle under dynamic operating conditions, and pressure data is combined with vibration data to analyze potential abnormal synchronization phenomena. Finally, through the above analysis, it is determined whether the vehicle is in an abnormal operating condition. In an abnormal state, a warning message is issued to the controller; in a normal operating condition, the above weight data, attitude data, pressure data, vibration dataset, and corresponding timestamps are recorded as the above operating condition information. The above technical solution effectively improves the weighing accuracy and can detect abnormal situations in real time when there is vibration, tilt, or other unstable factors. Through accurate detection of abnormal synchronization phenomena and real-time operating condition marking, the system can provide reliable data support for subsequent fault diagnosis, abnormal warning, and remote data interaction.

[0042] Step S2: Determine whether the high-precision weighing conditions are met based on the vehicle operating condition information; if the high-precision weighing conditions are met, lock the stabilization time window based on the determination result of whether the pressure data change trend has entered the stable range and the comparison result of the energy accumulation value of the vibration data with the preset threshold; within the stabilization time window, perform gravity component compensation, noise reduction processing, and outlier removal on the weight data to obtain an accurate weight value; specifically including:

[0043] Based on the vehicle operating condition information, it is determined whether the high-precision weighing conditions are met. When met, the rate of change between adjacent sampling points is calculated based on the pressure data, and it is determined whether the pressure data has entered a plateau period. Simultaneously, frequency domain analysis is performed on the vibration data, and the cumulative vibration energy value is calculated. The cumulative vibration energy value is compared with the preset threshold. When the pressure change trend is stable and the cumulative vibration energy value is lower than the preset threshold, the current time period is locked as a stable time window. Within the stable time window, multiple sets of weight data are continuously collected at a preset sampling frequency. Based on the corresponding attitude data, gravity component compensation processing is performed on the multiple sets of weight data. Noise reduction processing and outlier removal processing are performed on the compensated weight data. Statistical calculations are performed on the processed weight data to obtain the accurate weight value.

[0044] Specifically, based on the aforementioned vehicle operating condition information, it is determined whether the high-precision weighing conditions are met. These conditions include at least a stable pressure data trend, a stable tilt level, and a cumulative vibration energy value below a preset threshold. Specifically, the pressure data trend and the cumulative vibration energy value need to be analyzed to ensure accurate weighing under specific vehicle operating conditions. Based on the collected pressure data, the rate of change between adjacent sampling points is calculated, and this rate of change is used to determine if the pressure data has entered a plateau period. When the rate of change between several consecutive sampling points is consistently less than the preset stability threshold, the pressure trend is confirmed to be stable, indicating that the vehicle's load condition is stable. Furthermore, after the vibration data undergoes Fast Fourier Transform (FFT) processing, the time-domain signal is converted into frequency-domain data. The intensity of the vibration is further evaluated by calculating the cumulative vibration energy value and compared with a preset threshold obtained through experimental calibration. If the calculation result shows that the cumulative vibration energy value is lower than the preset threshold, it indicates that the vehicle is currently in a low-vibration state, meeting the high-precision weighing conditions.

[0045] Once the pressure data shows a stable trend and the cumulative energy of the vibration data is below a preset threshold, the current time period is designated as the stable time window. Within this window, multiple sets of weight data are continuously collected at a preset sampling frequency, typically 10 times per second, with at least five sets collected for subsequent processing. Gravity component compensation is then applied to the collected weight data. This process involves calculating the vehicle's tilt angle using attitude data and applying trigonometric formulas to compensate for the gravity component in the weight data based on the calculated tilt angle. The specific compensation process will be detailed later. The compensated weight data undergoes median filtering for noise reduction. The filter window size can be adjusted according to the specific application, typically set to 3 to 5 data points to effectively remove short-term random noise. Next, the mean and standard deviation of the processed data sequence are calculated. Outliers deviating from the mean by more than twice the standard deviation are removed to ensure data validity and accuracy. Finally, the average value is used as the precise weight value, resulting in a stable and reliable weighing result.

[0046] To further improve weighing accuracy, especially under complex working conditions such as significant vibration but relatively stable pressure, the preset threshold can be adjusted to 1.5 times to adapt to more complex environmental changes. The noise reduction filtering window size can also be increased to seven data points. This ensures accurate weight data is consistently provided under different road conditions and load conditions, supporting the efficient operation and monitoring of electric sanitation and food service vehicles. The above technical solution, through comprehensive analysis of the changing characteristics of pressure and vibration data, combined with gravity compensation, noise reduction, and outlier removal, ensures accurate vehicle load information even in complex environments. All data acquisition, processing, and calculation are performed within a specific time window, ensuring the accuracy of the weighing results and effectively addressing various interference factors under dynamic working conditions, thereby improving the system's reliability and applicability.

[0047] Furthermore, the acquisition of the gravity compensation component includes:

[0048] After locking the stable time window based on the pressure data change trend and the cumulative vibration energy value, the current tilt angle of the vehicle is calculated based on the attitude data. Then, according to the projection relationship of gravity on the sensor measurement direction, the weight data converted from the pressure signal is corrected for gravity components to obtain an initial compensation weight value. This initial compensation weight value is then time-aligned and normalized with the vibration characteristics, tilt angle characteristics, and pressure change characteristics within the corresponding time period. This result is then used as a multi-dimensional input feature vector and input to a pre-trained small neural network model. The small neural network model performs nonlinear prediction on the weight compensation residual and outputs a weight compensation correction amount. Finally, the weight compensation correction amount is superimposed on the initial compensation weight value to obtain the final weight compensation result.

[0049] Specifically, in order to improve the gravity compensation accuracy of the weight data within the stable time window and reduce the impact of vehicle micro-tilt, structural elastic rebound and slight vibration residuals on the weight compensation results, a gravity compensation acquisition process is executed after locking the stable time window in step S2. This process uses the initial compensation weight value obtained by physical projection correction as the basis, and uses a small neural network to perform secondary correction on the compensation residual as a fine compensation step, so that the final compensation result simultaneously satisfies physical interpretability and adaptive suppression capability against complex nonlinear errors.

[0050] During implementation, when the controller enters a plateau period based on the pressure data change trend and the cumulative vibration data energy value is lower than a preset threshold, it locks a stable time window. Within this stable time window, multiple sets of weight data are continuously collected at a preset sampling frequency, and corresponding attitude data are simultaneously collected to obtain the vehicle tilt angle. The weight data is preferably a weight measurement value obtained from a load cell, and the tilt angle reflects the degree of deviation of the gravity direction relative to the load cell's measurement axis. Subsequently, the controller performs gravity component correction on the weight data based on the projection relationship of gravity in the sensor's measurement direction. That is, the original weight measurement value is converted according to the projection coefficient corresponding to the tilt angle, so that the corrected weight value is equivalent to the vehicle being in a horizontal posture. The static load response at a given time is used as the initial compensation weight value. During this stage, the correspondence between weight data and attitude data is kept consistent through timestamp alignment, ensuring that each set of initial compensation weight values ​​has a corresponding tilt angle input, thereby guaranteeing the causal consistency and reproducibility of the compensation calculation. After obtaining the initial compensation weight value, the controller further introduces a small neural network model to perform a secondary correction on any remaining nonlinear errors from the initial compensation. To this end, the controller extracts multi-source feature data corresponding to each set of initial compensation weight values ​​within a stable time window and constructs an input feature vector. The defined multi-source feature data includes at least vibration signal features, tilt angle features, and pressure change features, and is time-stamped... The system ensures a one-to-one correspondence between the vibration signal characteristics and the initial compensation weight value. These characteristics can be represented by the energy, dominant frequency amplitude, or statistics of the vibration data within the window, reflecting the degree of influence of residual mechanical disturbance on the weighing structure. The tilt angle characteristics, in addition to the tilt angle value, can also include the tilt angle change rate or stability index within the window, representing the second-order effect of micro-attitude changes on gravity projection correction. The pressure change characteristics can be formed by the pressure change rate, plateau period fluctuation amplitude, etc., representing the minute fluctuations in the weighing stress state within the stability window. Subsequently, the above multi-source characteristics and the initial compensation weight value are normalized to eliminate dimensional differences and improve model inference stability. The normalized feature data are then spliced ​​together in a preset order. A multi-dimensional input feature vector is fed into a pre-trained small neural network model. The training data for this model comes from the initial compensation weight value and its corresponding vibration, attitude, and pressure change characteristics collected by the vehicle under known real load conditions within a stable time window. The difference between the precise weight value and the initial compensation weight value is used as a supervision label. By constructing a mapping relationship between the multi-dimensional input features and the compensation residual, forward propagation is used to calculate the predicted output residual value, and backpropagation is performed using the mean squared error as the loss function to update the network weights. After multiple rounds of iterative training until the loss converges, model parameters capable of nonlinearly predicting gravity compensation residuals are obtained. The correction effect is as follows: Figure 2 As shown.

[0051] Step S3: When the high-precision weighing conditions are not met, analyze the attitude data, weight data, vibration data, and pressure data within the current time period, and match them with a pre-built error model library to obtain the corresponding error type; assign corresponding weights to the attitude data, vibration data, and pressure data according to the error type, and obtain the corresponding weight estimate value by combining it with the weight estimation model; specifically including:

[0052] When high-precision weighing conditions are not met, attitude data, weight data, vibration data, and pressure data are collected in real time within the current time period. The tilt angle change rate, vibration energy level, pressure change rate, and their time synchronization characteristics are extracted to construct a dynamic working condition feature vector. This dynamic working condition feature vector is matched with a preset dynamic working condition mapping relationship model library. Based on the matching results, multidimensional interpolation is used to calculate the fluctuation range and offset trend of the weight data under the current dynamic working condition. The mapping relationship model library is established hierarchically based on historical calibration data, categorized by vibration level, tilt angle range, and load range. Dynamic correction coefficients for attitude data, vibration data, and pressure data are calculated based on the fluctuation range and offset trend of the weight data. These coefficients are then normalized and their change rate constrained to form corresponding dynamic weights. Within a preset sliding time window, the real-time acquired attitude data, vibration data, pressure data, and corresponding dynamic weights are input into the weight estimation model for fusion calculation. The output results are then time-smoothed to obtain the weight estimate under the current dynamic working condition.

[0053] Specifically, to address the issue of fluctuating and drifting weight results and insufficient usability of electric sanitation and kitchen vehicle weighing vehicles under complex dynamic conditions, step S3 no longer simply performs gravity compensation and smoothing on the weight data. Instead, when any situation occurs that prevents high-precision weighing from being met, such as the pressure change trend not reaching a stable plateau, the accumulated vibration energy exceeding a preset threshold, or the vehicle being tilted, the process switches to weight estimation under dynamic conditions. At a preset sampling period, posture data, vibration data, and pressure data are simultaneously collected within the current time period, and timestamps are aligned between multiple data sources to ensure comparability and fusion of data from different sensors within the same sliding window. Subsequently, after basic denoising and dimensional unification processing, the tilt angle value is calculated from the posture data, and the tilt angle change is further determined. The rate of change of tilt angle is used to characterize the intensity of dynamic changes in vehicle attitude. The tilt angle change rate is used to reflect whether the vehicle tilt state is in a rapid change process. Vibration energy index is obtained from vibration data through time-domain to frequency-domain conversion and vibration energy level is formed according to preset classification rules to characterize the level of mechanical disturbance to the vehicle. At the same time, the pressure change rate between adjacent sampling points is calculated from pressure data and pressure change characteristics are obtained by combining the pressure peak occurrence time. Based on this, the controller combines the tilt angle change rate, vibration energy level, pressure change rate and the time synchronization characteristics between the three to construct a dynamic working condition feature vector. The time synchronization characteristics are used to reflect the degree of synchronization between vibration peak and pressure peak within the sliding window, thereby distinguishing between "real load gradual change" and "transient force fluctuation caused by vibration impact" to avoid mistaking impact component for load change.

[0054] After obtaining the dynamic operating condition feature vector, the controller calls the preset dynamic operating condition mapping relationship model library and performs matching calculations. This dynamic operating condition mapping relationship model library is preferably constructed from traceable data from the vehicle calibration phase or historical operation phase, and is organized hierarchically according to vibration energy level, tilt angle range, and load range, so that each layer corresponds to a set of mapping relationships between "pressure change characteristics and weight fluctuation patterns" under known operating conditions. Specifically, during construction, the precise weight value obtained within a stable time window that meets high-precision weighing conditions can be used as a reference benchmark. In adjacent unstable windows, the fluctuation amplitude and offset trend of the weight data relative to the reference benchmark are statistically analyzed, thereby establishing a correspondence between the "operating condition feature vector" and the "weight fluctuation range and offset trend" and storing it in the model. A model library is established to ensure that the mapping relationships in the model library have clear physical meaning and reproducible engineering sources. Based on the above-mentioned hierarchical model library structure, the controller calculates the similarity between the feature vector and the candidate working condition vector in the model library under the current working condition, and selects at least two adjacent working conditions with the highest similarity as the interpolation benchmark. Then, multi-dimensional interpolation calculation is performed based on the relative position of the current feature vector between adjacent working conditions to obtain the fluctuation range and offset trend of the weight data under the current dynamic working condition. The fluctuation range is used to limit the upper bound of the short-term disturbance amplitude introduced by vibration and attitude changes, and the offset trend is used to describe the long-term drift direction and speed of the estimated value caused by the slow change of pressure and attitude, so that the subsequent estimation process can simultaneously take into account "suppressing instantaneous jitter" and "tracking slow drift".

[0055] After obtaining the fluctuation range and offset trend, the controller further calculates dynamic correction coefficients for attitude data, vibration data, and pressure data, and explicitly uses these correction coefficients to form dynamic weights, reflecting the contribution of different data sources to the reliability of weight estimation under the current operating conditions. Specifically, the attitude correction coefficient reflects the influence of tilt angle changes on gravity projection errors in weight measurement; preferably, its gain on instantaneous weight updates decreases with increasing tilt angle change rate to avoid miscompensation caused by rapid attitude swaying. The vibration correction coefficient reflects the degree of contamination of weight data fluctuations by vibration disturbances; preferably, its suppression intensity increases with increasing vibration energy level to reduce the high-frequency response of weight data under high vibration. The pressure correction coefficient reflects the influence of pressure change rate on the stability of weight estimation; preferably, it combines the offset trend to constrain the update speed of the estimated value to avoid sudden changes in estimation caused by transient pressure shocks. To ensure the stability of the estimation process, the controller performs normalization processing on the above correction coefficients to satisfy the constraint that the weight sum is a preset constant, and further applies a change rate constraint to each dynamic correction coefficient, ensuring that the change in weight between adjacent sliding windows does not exceed a preset upper limit, thereby avoiding oscillations in the estimated value caused by drastic weight jumps due to short-term abnormal pulses.

[0056] Subsequently, within a preset sliding time window, the controller inputs time-aligned attitude data, vibration data, pressure data, and their corresponding dynamic weights into the weight estimation model for fusion calculation. This weight estimation model can be implemented using a weighted fusion state update structure, where the instantaneous weight converted from pressure is used as the observation, the weight component after attitude compensation is used as the geometric correction term, the vibration suppression term is used as the observation noise gain adjustment factor, and the pressure trend constraint term is used as the limiting condition for state changes. This ensures that the model output is more biased towards "steady-state maintenance" during high-disturbance phases and gradually returns to "tracking updates" during disturbance reduction phases. During the fusion process, the controller performs threshold checks on instantaneous observation deviations based on the fluctuation range and removes outliers exceeding the upper limit of the fluctuation to prevent impact peaks from directly entering the fusion calculation. Simultaneously, it applies gradual correction to the state term based on the offset trend, ensuring that the estimated value remains continuous under unstable conditions and has a directionality consistent with the actual load change. Finally, the controller... The output of the weight estimation model is subjected to time smoothing. The time smoothing is preferably implemented by moving average or first-order low-pass filtering, and can be combined with a limiting mechanism to constrain the maximum change per unit time within a preset range, thereby obtaining the weight estimate under the current dynamic working condition and outputting it for local storage and subsequent upload scheduling. Through the above technical solution, when the high-precision weighing conditions are not met, a closed-loop link of "dynamic working condition feature vector, model library mapping, interpolation to obtain fluctuation and drift, weight-constrained dynamic allocation, sliding window fusion and time smoothing output" is used to achieve the orderly utilization of multi-source information. This allows attitude data, vibration data and pressure data to correspond to different processing paths of "geometric error correction", "disturbance contamination suppression" and "slow drift constraint", respectively, and to be uniformly fused in the same estimation model through dynamic weights. This avoids simply stacking terms without abstract description of data relationships, and significantly improves the stability and usability of weight estimation under complex dynamic working conditions in engineering.

[0057] Furthermore, step S3 also includes:

[0058] After obtaining the weight estimate, the most recent accurate weight value obtained within a stable time window is retrieved. The difference and relative error between the weight estimate and the accurate weight value are calculated to obtain a comparison result. When the difference or relative error is within a preset error range, the parameters in the fusion estimation algorithm are corrected using the accurate weight value as a calibration benchmark. The parameters include at least one of attitude correction coefficient, vibration suppression coefficient, pressure trend constraint coefficient, or fusion weight. The corresponding parameters are adjusted by scaling or bias compensation to make subsequent weight estimates converge to the accurate weight value. When the difference exceeds the preset error range, the current fusion estimation algorithm parameters are kept unchanged or the parameter update magnitude is reduced.

[0059] Specifically, to ensure that the weight estimate obtained in step S3 remains consistent with the actual load under long-term dynamic operating conditions, and to prevent the parameters of the fusion estimation algorithm from accumulating deviations due to environmental changes, a self-correction process based on the accurate weight value is further executed after the weight estimate calculation is completed. Specifically, when the high-precision weighing conditions are met again during vehicle operation and a stable time window is locked, the accurate weight value is obtained according to step S2 and stored together with the corresponding timestamp; subsequently, after the weight estimate is output in step S3, the closest and most valid accurate weight value in time is retrieved from the local storage module based on the timestamp as a reference benchmark value; the difference between the current weight estimate and the above accurate weight value is calculated to obtain the absolute error. The relative error between the two is calculated, which is obtained by dividing the difference by the precise weight value to eliminate the scale influence of different load ranges on the error judgment. Based on this, a comparison result is formed, and the above difference and relative error are compared with a preset error range. The preset error range can be set according to the vehicle load level. For example, a smaller error threshold is allowed in the light load range, and a moderate relaxation is allowed in the heavy load range, so as to ensure that the correction strategy has regional adaptability. When the comparison result shows that the difference or relative error is within the corresponding preset error range, it is considered that there is a correctable systematic deviation between the current fusion estimation algorithm output and the actual load, rather than a sudden anomaly or disturbance caused by changes in the actual load, and the parameter correction mechanism is activated.

[0060] The aforementioned parameter correction mechanism uses the precise weight value as a calibration benchmark to adjust key parameters in the fusion estimation algorithm. These key parameters include at least one of the following: attitude correction coefficient, vibration suppression coefficient, pressure trend constraint coefficient, and fusion weight. Specifically, based on the error direction, it determines whether the estimated value tends to be overestimated or underestimated. A scaling factor or bias compensation is calculated proportionally based on the error magnitude, thereby performing proportional scaling or incremental correction on the corresponding parameters. For example, when the overall estimated value is too high but the error is within the allowable range, the vibration suppression coefficient is appropriately reduced or the proportion of the pressure component in the fusion weight is adjusted to make the subsequent estimation process more conservative in response to transient pressure changes. When the overall estimated value is too low, the attitude correction coefficient is correspondingly increased or the pressure trend constraint coefficient is adjusted to enhance the tracking ability of the fusion estimation algorithm to real load changes while maintaining stability. To prevent system oscillations caused by parameter adjustments, the controller sets a maximum update ratio for each correction magnitude. The system employs a limiting and decreasing update factor to ensure that parameter adjustments gradually converge, thereby guaranteeing that subsequent weight estimates gradually approach the accurate weight value without overshoot. When the comparison results show a difference or relative error exceeding the aforementioned preset error range, it is determined that the current difference may be caused by instantaneous abnormal operating conditions, short-term sensor changes, or changes in the actual load. To avoid erroneous data affecting the fusion estimation algorithm's correction, the controller maintains the current fusion estimation algorithm parameters unchanged or reduces the parameter update amplitude to a preset lower limit, and records the number of anomalies for subsequent diagnostic analysis. This technical solution, through periodic comparison based on accurate weight values ​​and a controlled parameter correction mechanism, enables the dynamic fusion estimation algorithm in step S3 to maintain real-time estimation capabilities under complex operating conditions, while also allowing for adaptive calibration using high-precision results as anchors when conditions permit. This suppresses error accumulation during long-term operation and improves the stability, consistency, and engineering reliability of the weight estimation results.

[0061] Step S4: Store the precise weight value or the estimated weight value, and upload it to the remote server according to the upload order determined by the preset task priority rules;

[0062] The precise weight value or the estimated weight value is stored by an upper-mounted information controller; the data upload order is determined according to a hard task priority matrix, which includes at least security alarm tasks, fault diagnosis tasks, operation data tasks, and remote upgrade tasks; the security alarm tasks maintain the highest priority under different network conditions.

[0063] Specifically, after obtaining the accurate weight value or weight estimate, in order to ensure the reliable preservation and orderly uploading of the data, further data storage and uploading control is performed. That is, the above-mentioned accurate weight value or weight estimate is uniformly managed by the upper-mounted information controller, and the data is uploaded to the remote server according to the preset task priority rules to determine the data upload order. In this way, a balance is achieved between data security, real-time performance and overall system efficiency in complex network environments and multi-task concurrent scenarios.

[0064] The precise weight value or weight estimate is stored in an internal storage module via an upper-mounted information controller. This storage module preferably uses non-volatile storage media, such as flash memory or other power-off retention memory, to ensure that the weighing data is not lost in the event of vehicle shutdown, power failure, or communication interruption. During storage, not only the weight value itself is preserved, but the weight data is also associated with a corresponding timestamp and the vehicle's unique identification information, thus forming a traceable data record with complete time-series information and vehicle ownership. This structured storage method enables accurate reconstruction of the vehicle's condition during subsequent local queries, anomaly retrospective analysis, or cloud-based analysis. The system tracks the operational status and load at specific time points, and reduces storage latency through simplified data writing logic, improving its responsiveness to real-time weighing data. After local storage, instead of directly sending all data to be uploaded to the remote server in the order of generation, a hard task priority matrix is ​​introduced to uniformly schedule and manage various upload tasks, including weight data. This hard task priority matrix is ​​a pre-defined task scheduling rule table used to quantify and sort different types of data upload tasks. It includes at least safety alarm tasks, fault diagnosis tasks, operational data tasks, and remote upgrade tasks. Among these, safety alarm tasks are assigned the highest priority score, which remains the highest in the matrix. Fault diagnosis tasks are next, operational data tasks (including the upload of precise weight values ​​or weight estimates) have a medium priority, and remote upgrade tasks have a relatively low priority.

[0065] In actual operation, the upper-mounted information controller monitors the current network status in real time and classifies it into different levels, such as good, weak signal, and network outage. When multiple tasks are pending upload, the system sorts them according to their priority scores in the hard task priority matrix and adds them to the upload queue in sequence, ensuring that high-priority tasks occupy communication resources first. For example, in situations where network bandwidth is limited or the signal is weak, the system can prevent low-priority tasks from occupying limited bandwidth and blocking critical data transmission, allowing safety alarm tasks and fault diagnosis tasks to be uploaded first. Weight data, as a type of operational data task, is uploaded to the remote server in sequence after the upload requirements of high-priority tasks are met. In particular, to ensure vehicle operation safety, this invention explicitly stipulates that safety alarm tasks always maintain the highest priority under different network conditions. Even in scenarios with weak signals or network outages, when a safety-related anomaly is detected and alarm data is generated, the alarm data is immediately placed at the top of the upload queue. If the current network is completely interrupted, the alarm data and weight data are cached together in the local storage module, and the alarm data is sent first after the network is restored, followed by the upload of operational data such as weight values ​​according to priority. Through this mechanism, weight data is neither lost in the overall data link nor interfered with the real-time uploading of safety-related data, thus forming a complete closed-loop process from weighing result generation, local data storage to orderly uploading to the cloud. Through the above-mentioned storage and priority upload control mechanism, not only is the integrity and traceability of weighing data from electric sanitation and kitchen vehicles guaranteed under complex operating conditions and changing communication environments, but also, through the introduction of a task priority matrix, intelligent scheduling of multiple types of data is achieved, effectively improving the system's comprehensive performance in terms of safety response, operation monitoring, and remote management, making the information-based weighing system more in line with the needs of actual engineering applications.

[0066] This invention also provides an information-based weighing system control system for electric sanitation and kitchen waste trucks, used to implement the above-mentioned methods, such as... Figure 3 As shown, the system includes:

[0067] The data acquisition unit is used to collect weight data, posture data, vibration data, and pressure data of the electric sanitation and food processing vehicle during operation; and to determine the current vehicle operating condition information based on the weight data and posture data.

[0068] The precision measurement unit is used to determine whether the high-precision weighing conditions are met based on the vehicle operating condition information; if the high-precision weighing conditions are met, the unit locks the stable time window based on the determination result of whether the change trend of the pressure data has entered the stable range and the comparison result of the energy accumulation value of the vibration data with the preset threshold; within the stable time window, the weight data is compensated for the gravity component to obtain the accurate weight value.

[0069] The weighing estimation unit is used to obtain the corresponding error type by matching the attitude data, vibration data, and pressure data in the current time period with a pre-built error model library when the high-precision weighing conditions are not met; to assign corresponding weights to the attitude data, vibration data, and pressure data according to the error type, and to obtain the corresponding weight estimate value by combining the fusion estimation algorithm.

[0070] The data storage and uploading unit is used to store the precise weight value or the weight estimate, and upload it to the remote server after determining the upload order according to the preset task priority rules.

[0071] The present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the above-described method.

[0072] In summary, this invention, by simultaneously collecting multi-source data such as weight, attitude, vibration, and pressure during vehicle operation, comprehensively analyzes the vehicle's load state, tilt state, and vibration disturbance characteristics to achieve refined identification of the current operating condition. Furthermore, by using the temporal synchronization of vibration and pressure, it identifies potential abnormal operating conditions, providing reliable criteria for subsequent weighing mode selection. This establishes a basis for separating stable and unstable operating conditions, avoiding the blind output of high-precision results when conditions are not met. When high-precision weighing conditions are met, a stable time window is locked by jointly determining the pressure change trend and vibration energy threshold. Within this window, gravity component compensation, noise reduction, and outlier removal are implemented. A small neural network is then used to nonlinearly correct residual errors, thereby obtaining accurate weight values ​​close to static weighing effects. This is achieved through a collaborative compensation mechanism between the physical model and the data-driven model. This method improves measurement accuracy under slightly disturbed environments. Instead of simply abandoning weighing when high precision conditions are not met, it matches dynamic working condition feature vectors with a pre-defined mapping model library, obtains fluctuation range and offset trends through multi-dimensional interpolation, and outputs weight estimates using a dynamic weight fusion estimation algorithm. This enables continuous tracking of the weighing status under complex dynamic working conditions. Furthermore, by comparing the estimated weight with accurate weight values ​​within subsequent stable windows, the fusion parameters are dynamically corrected, gradually converging the estimation results and preventing long-term error accumulation. Finally, through local data storage and task priority scheduling mechanisms, the method ensures orderly and secure uploading of weighing data under different network environments, prioritizing the transmission of safety alarm information and ensuring stable feedback of operational data. This method achieves high-precision, robust, and highly reliable load information management in complex operating environments of electric sanitation and catering vehicles.

[0073] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0074] 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, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0075] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention 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 the present invention.

Claims

1. A control method for an information-based weighing system of an electric sanitation and kitchen waste vehicle, characterized in that, The method includes: Step S1: Collect weight data, posture data, vibration data, and pressure data of the electric sanitation and kitchen waste vehicle during operation; determine the current vehicle operating condition information based on the weight data and posture data; Step S2: Determine whether the high-precision weighing conditions are met based on the vehicle operating condition information; if the high-precision weighing conditions are met, lock the stable time window based on the determination result of whether the change trend of the pressure data has entered the stable range and the comparison result of the energy accumulation value of the vibration data with the preset threshold; within the stable time window, perform gravity component compensation, noise reduction processing and outlier removal on the weight data to obtain the accurate weight value. Step S3: When the high-precision weighing conditions are not met, the attitude data, vibration data, and pressure data in the current time period are matched with the pre-built error model library to obtain the corresponding error type; the attitude data, vibration data, and pressure data are assigned corresponding weights according to the error type, and the corresponding weight estimate is obtained by combining the fusion estimation algorithm. Step S4: Store the precise weight value or the weight estimate, and upload it to the remote server according to the upload order determined by the preset task priority rules.

2. The method as described in claim 1, characterized in that, In step S1, the current vehicle operating condition information is determined, including: The system calculates the ratio of the average weight data within a set time period to the vehicle's unloaded weight, and determines the load status as heavy or light based on a first comparison result between the ratio and a preset ratio threshold. It also calculates the pressure data variation amplitude within the set time period and compares it with a preset amplitude to obtain a second comparison result, determining the degree of swaying as bumpy or stable based on the second comparison result. The system performs a Fast Fourier Transform on the vibration data to obtain the dominant vibration frequency band information, aligns the peak value of the dominant vibration frequency band information with the peak value of the pressure data according to timestamps, calculates the degree of synchronization between the two within a preset sliding window, and compares it with a synchronization judgment threshold to determine if there is an abnormal synchronization phenomenon between vibration and pressure. When an abnormal synchronization phenomenon between vibration and pressure exists, the current time is marked as a potential abnormal operating condition point, and the weight data, attitude data, vibration data, pressure data, and corresponding timestamps corresponding to non-potential abnormal operating condition points are used as the vehicle operating condition information.

3. The method according to claim 1, characterized in that, In step S2, the accurate weight value is obtained including: Based on the vehicle operating condition information, it is determined whether the high-precision weighing conditions are met. These conditions include at least a stable pressure data trend and a cumulative vibration energy value below a preset threshold. When met, the rate of change between adjacent sampling points is calculated based on the pressure data, and it is determined whether the pressure data has entered a plateau period. Simultaneously, frequency domain analysis is performed on the vibration data, and the cumulative vibration energy value is calculated. The cumulative vibration energy value is compared with the preset threshold. When the pressure trend is stable and the cumulative vibration energy value is below the preset threshold, the current time period is locked as a stable time window. Within the stable time window, multiple sets of weight data are continuously collected at a preset sampling frequency. Gravity component compensation processing is performed on the multiple sets of weight data based on the corresponding attitude data. Noise reduction and outlier removal processing are performed on the compensated weight data. Statistical calculations are then performed on the processed weight data to obtain the accurate weight value.

4. The method according to claim 3, characterized in that, In step S2, the acquisition of the gravity compensation component includes: After locking in a stable time window based on the pressure data change trend and the cumulative vibration energy value, the vehicle's current tilt angle is calculated based on the attitude data. Then, based on the projection relationship of gravity in the sensor measurement direction, the original weight measurement value obtained from the pressure signal is corrected for gravity component to obtain an initial compensation weight value. This initial compensation weight value is then time-aligned and normalized with the vibration signal characteristics, tilt angle characteristics, and pressure change characteristics within the corresponding time period. This is then used as a multi-dimensional input feature vector and input to a pre-trained small neural network model. The small neural network model performs nonlinear prediction on the weight compensation residual and outputs a weight compensation correction amount. Finally, this weight compensation correction amount is superimposed on the initial compensation weight value to obtain the final weight compensation component.

5. The method according to claim 1, characterized in that, In step S3, when the high-precision weighing conditions are not met, the corresponding weight estimate is obtained, including: When high-precision weighing conditions are not met, attitude data, weight data, vibration data, and pressure data are collected in real time within the current time period. The tilt angle change rate, vibration energy level, pressure change rate, and their time synchronization characteristics are extracted to construct a dynamic working condition feature vector. This dynamic working condition feature vector is matched with a preset dynamic working condition mapping relationship model library. Based on the matching results, multidimensional interpolation is used to calculate the fluctuation range and offset trend of the weight data under the current dynamic working condition. The mapping relationship model library is established hierarchically based on historical calibration data, categorized by vibration level, tilt angle range, and load range. Dynamic correction coefficients for attitude data, vibration data, and pressure data are calculated based on the fluctuation range and offset trend of the weight data. These coefficients are then normalized and their change rate constrained to form corresponding dynamic weights. Within a preset sliding time window, the real-time acquired attitude data, vibration data, pressure data, and corresponding dynamic weights are input into the weight estimation model for fusion calculation. The output results are then time-smoothed to obtain the weight estimate under the current dynamic working condition.

6. The method according to claim 1, characterized in that, The construction process of the mapping relationship model library includes: By operating the vehicle under different vibration levels, tilt angle ranges, and load ranges, and simultaneously collecting attitude data, vibration data, and pressure data, the fluctuation amplitude and offset of the weight measurement value converted from pressure relative to the actual weight value are calculated within the corresponding unstable time period. Operating condition characteristic parameters, including at least the tilt angle change rate, vibration energy level, and pressure change rate, are extracted. A one-to-one correspondence is established between these operating condition characteristic parameters and the corresponding weight fluctuation amplitude and offset, and the data is stratified and categorized according to vibration level, tilt angle range, and load range to form a multidimensional mapping data table. The multidimensional mapping data table is then interpolated and smoothed to generate a continuously queryable mapping relationship model library.

7. The method according to claim 1, characterized in that, Step S3 also includes: After obtaining the weight estimate, the most recent accurate weight value obtained within a stable time window is retrieved. The difference and relative error between the weight estimate and the accurate weight value are calculated to obtain a comparison result. When the difference or relative error is within a preset error range, the parameters in the fusion estimation algorithm are corrected using the accurate weight value as a calibration benchmark. The parameters include at least one of attitude correction coefficient, vibration suppression coefficient, pressure trend constraint coefficient, or fusion weight. The corresponding parameters are adjusted by scaling or bias compensation to make subsequent weight estimates converge to the accurate weight value. When the difference exceeds the preset error range, the current fusion estimation algorithm parameters are kept unchanged or the parameter update magnitude is reduced.

8. A control system for an information-based weighing system of an electric sanitation and kitchen waste vehicle, used to implement the method as described in any one of claims 1-7, characterized in that, The system includes: The data acquisition unit is used to collect weight data, posture data, vibration data, and pressure data of the electric sanitation and food processing vehicle during operation; and to determine the current vehicle operating condition information based on the weight data and posture data. The precision measurement unit is used to determine whether the high-precision weighing conditions are met based on the vehicle operating condition information. If the high-precision weighing conditions are met, the unit locks the stable time window based on the determination result of whether the change trend of the pressure data has entered the stable range and the comparison result of the energy accumulation value of the vibration data with the preset threshold. Within the stable time window, the weight data is subjected to gravity component compensation, noise reduction processing and outlier removal to obtain an accurate weight value. A simplified weighing unit is used to obtain a weight estimate by using a simplified processing method if the high-precision weighing conditions are not met. The data storage and uploading unit is used to store the precise weight value or the weight estimate, and upload it to the remote server after determining the upload order according to the preset task priority rules.

9. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the method as described in any one of claims 1-7.

Images