An adaptive adjustment method for tomato transportation speed based on path condition feedback

Abstract

The present invention relates to the field of intelligent transport control technology, specifically to a method for adaptively adjusting tomato transport speed based on route condition feedback. The method comprises: using a multi-source sensing unit to collect road images, vibration waveforms, and altitude data to construct real-time road condition information; identifying driving modes and extracting corresponding bump parameters; combining multispectral and thermal imaging to detect tomato maturity levels, and then querying a maturity-compressive strength correspondence table to calculate cargo damage thresholds; constructing a dynamic mapping model under multiple operating conditions to predict vibration responses and convert them into equivalent pressures; comparing the equivalent pressures with the damage thresholds to obtain a safety margin, constructing a speed adjustment decision tree, and generating maximum allowable speed values ​​for each road section; and dynamically generating segmented speed change control commands based on the speed decision matrix to control coordinated speed changes between the throttle and the braking system. This method offers advantages such as accurate operating condition identification, dynamic fruit adaptation, and a closed-loop speed control system, making it suitable for precise speed control in highly sensitive fruit and vegetable transport scenarios.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation control technology, and in particular to a method for adaptively adjusting tomato transportation speed based on path condition feedback. Background Art

[0002] As a high-value and fragile fruit and vegetable, tomatoes experience a significant decrease in compressive strength of their peel in the late stages of maturity. These fruits are easily damaged by road bumps, mechanical shock, or loading and lamination during transportation, which can affect product quality and economic benefits. Especially in multi-path, multi-condition transportation scenarios, where vehicles often need to traverse different types of roads (such as asphalt roads, cement roads, and rail sections) and complex gradient environments, the uncontrollability of vibration and impact during transportation increases significantly. To improve the in-transit integrity of fruits and ensure the quality of flexible logistics transportation, the industry has attempted in recent years to integrate route planning, speed control, and vibration monitoring into the transportation process to achieve a dynamic match between the transportation rhythm and the cargo's tolerance.

[0003] However, existing technologies often focus on developing static transport speed control strategies based on speed limits or vehicle dynamic performance, lacking the ability to comprehensively model real-time changes in road conditions, driving mode transitions, and the physical properties of fruit (such as changes in maturity and compressive strength). Most methods are unable to decouple the operating conditions of land and rail transport, nor do they establish a physical mapping mechanism for vibration parameters to cargo stress responses. This results in overly coarse transport speed control strategies, delayed control, or distorted responses. Furthermore, traditional solutions generally lack a precise speed segmentation control mechanism for risky sections of the transport route, an adaptive feedback mechanism, and command reconstruction capabilities, making them unable to meet the safety requirements of "flexible transportation" for highly sensitive fruit. Summary of the Invention

[0004] The present invention provides a method for adaptively adjusting the tomato transportation speed based on path condition feedback, and provides an intelligent control method that integrates multi-source perception, fruit characteristic adaptation and path dynamic speed regulation capabilities.

[0005] A method for adaptively adjusting tomato transportation speed based on path condition feedback includes the following steps:

[0006] S1: Acquire real-time road condition information of the target transport route, wherein the road condition information includes road type, road surface flatness, and altitude gradient;

[0007] S2: Identify the current driving mode based on real-time road condition information. Driving modes include track mode and land mode, and calculate corresponding bump parameters, including vertical vibration frequency, lateral swing amplitude and duration in land mode, track joint impact strength and upper and lower rail transition section inclination angle in track mode, as well as maximum impact acceleration and attitude adjustment time when switching modes.

[0008] S3: Based on the tomato maturity detection data and the preset maturity-compressive strength correspondence table, the damage threshold of the tomatoes in the current cargo box is calculated;

[0009] S4: Establish a dynamic mapping model between cargo box vibration amplitude and bump parameters, distinguish between track and land conditions and generate vibration response prediction curves;

[0010] S5: Determine the maximum permissible speed value for each road section based on the difference between the vibration response prediction curve and the damage threshold;

[0011] S6: Generate a segmented speed change control command based on the maximum allowable speed value and perform a pre-deceleration operation before entering the target section.

[0012] Optionally, the S1 includes:

[0013] S11: Synchronously collecting physical feature data of the target transportation path through a multi-source sensing unit, wherein the physical feature data includes a road image sequence captured by a machine vision module, a road surface vibration waveform measured by a three-axis accelerometer, and an altitude point set recorded by an RTK-GPS module;

[0014] S12: Performing convolutional neural network classification processing on the road image sequence to output structured road type parameters, wherein the road type parameters include classification codes and corresponding confidence levels of asphalt pavement, cement pavement, and gravel pavement;

[0015] S13: Performing frequency domain energy analysis on the road surface vibration waveform, extracting the vibration root mean square value in the 5-20 Hz frequency band, and converting it into a standard road surface roughness index using a calibration curve, wherein the road surface roughness index is in mm / m;

[0016] S14: performing a sliding window difference calculation on the elevation point set to generate a continuous elevation change rate curve, and extracting road section data with an absolute value of a change rate greater than 3% in the curve as an effective elevation change gradient parameter;

[0017] S15: The road type parameter, road surface roughness index, and altitude change gradient parameter are integrated according to a unified spatiotemporal reference to generate a real-time road condition information data packet with a timestamp, wherein the road type, road surface roughness, and altitude change gradient parameters are updated at a frequency of 10 Hz per second.

[0018] Optionally, the S2 includes:

[0019] S21: Execute mode decision logic based on the road type parameter in the real-time road condition information generated in S15: if the road type is a railway track, trigger the track mode; otherwise, trigger the land mode;

[0020] S22: Calculate the bump parameters in the land mode and calculate the corresponding vertical vibration frequency based on the road surface roughness index provided by S15;

[0021] Using the altitude gradient parameter in S15 and the vehicle wheelbase, the lateral swing amplitude of the vehicle under the slope is calculated;

[0022] Identify road sections with a continuous slope of 3% or more, and deduce the duration of the vehicle's action on this type of slope based on the current vehicle speed, for subsequent speed adjustment decisions.

[0023] S23: performing bump parameter calculation in track mode: calculating the wheel-rail impact cycle based on the fixed track joint spacing and the current vehicle speed, and extracting the acceleration peak within the cycle range as the track joint impact intensity;

[0024] At the same time, the elevation difference between adjacent track segments is extracted, and combined with the track segment length, the inclination angle of the upper and lower track transition section is calculated to evaluate the posture change trend of the vehicle in track mode.

[0025] Optionally, the S2 further includes:

[0026] S24: If a mode switch occurs during transportation (i.e., switching from land mode to track mode, or vice versa), the maximum acceleration value at the switching moment is captured by the acceleration sensor and recorded as the maximum impact acceleration;

[0027] Secondly, the time from when the switching command is issued to when the vehicle's attitude angle stabilizes within a range of plus or minus 1 degree is tracked. This time is the attitude adjustment time and is used to evaluate the transition impact caused by mode switching.

[0028] S25: The turbulence parameters calculated in S22, S23, and S24 are uniformly encapsulated into a structured turbulence parameter data package. The data package includes: vertical vibration frequency, lateral swing amplitude, and duration of action in the land mode parameters; track joint impact strength and inclination angle of the upper and lower rail transition sections in the track mode parameters; and maximum impact acceleration and attitude adjustment time in the mode switching parameters.

[0029] Optionally, the S3 includes:

[0030] S31: The multispectral imaging unit on top of the cargo box scans the surface of the tomatoes, collecting the dual-band reflectance ratio of 520nm and 680nm wavelengths. This is combined with infrared thermal image information to identify the maturity of each tomato.

[0031] Generate tomato maturity detection data based on preset image recognition logic. The tomato maturity detection data includes the maturity grade of each tomato, which is divided into grades 1 to 5. At the same time, a spatial distribution matrix reflecting the spatial location of tomatoes at each grade is generated for subsequent processing;

[0032] S32: calling a pre-stored maturity-compressive strength correspondence table, where the maturity-compressive strength correspondence table presents compressive strength values ​​corresponding to different maturity levels in a two-dimensional array format.

[0033] According to the tomato maturity level results output in S31, the maturity-compressive strength correspondence table is called one by one, and the maturity level of each tomato is used as the table lookup index. The compressive strength values ​​corresponding to all tomatoes are batch matched and extracted as the basic input for subsequent calculations.

[0034] S33: Combining all the compressive strength values ​​in the tomato spatial distribution matrix, extracting the minimum compressive strength value as a baseline compressive strength value. Based on a preset damage threshold calculation rule and the number of stacked layers of tomatoes in the container, the baseline compressive strength value is corrected to obtain a damage threshold for the tomatoes in the container. This threshold is used to assess the maximum tolerance of the tomatoes to vibration and shock during transportation.

[0035] S34: The calculated damage threshold and the jolt parameter data output in S25 are synchronously packaged into a cargo state feature vector. This feature vector fully describes the load tolerance characteristics of the tomatoes in the current transport state and serves as an important input for vibration modeling and speed adjustment control in the subsequent step S4.

[0036] Optionally, the S4 includes:

[0037] S41: parsing the bump parameters output by S25, and classifying the parameter contents into three independent data sets according to the mode identification fields therein;

[0038] The land mode data set includes three indicators: vertical vibration frequency, lateral oscillation amplitude and duration of action;

[0039] The track pattern dataset includes the impact strength of track joints and the inclination angle of the transition section between the upper and lower rails;

[0040] The mode switching data set includes the maximum impact acceleration and attitude adjustment time;

[0041] The purpose of this step is to provide a clear data basis for modeling different working conditions separately;

[0042] S42: Based on data grouping, a dynamic mapping model with decoupled working conditions is constructed. The dynamic mapping model consists of three sub-models, corresponding to the land mode, track mode and mode switching scenario respectively.

[0043] The land mode model predicts the vibration response of the vehicle in this state through vertical vibration frequency, lateral swing amplitude and duration;

[0044] The track mode model describes the response to impact accumulation through the impact strength of the track joint and the inclination angle of the transition section between the upper and lower rails;

[0045] The mode switching model describes the transient vibration changes generated when the vehicle transitions from one mode to another based on the maximum impact acceleration and the posture adjustment time.

[0046] The parameters of all models are obtained through prior experiments and are constructed for different working conditions to enhance their adaptability to actual transportation conditions.

[0047] S43: Input the various data sets separated in S41 into the corresponding dynamic mapping models to generate vibration response prediction curves under land conditions, track conditions, and mode switching conditions, respectively.

[0048] The output of the land working condition is the vibration amplitude-time curve covering the frequency range of 0 to 20 Hz;

[0049] Track working condition outputs the vibration peak sequence of the track joint impact point;

[0050] Transient vibration envelope within 3 seconds after the output is switched in the mode switching condition;

[0051] All prediction curves are presented in the form of high-frequency time domain data, and the sampling rate is set to 1000 data points per second to meet the needs of fine-grained response analysis.

[0052] S44: performing a model validation and optimization process on the generated vibration response prediction curve, comparing the prediction curve with the measured data obtained by the vehicle accelerometer during the actual transportation process point by point. If the relative error within a certain period exceeds a set threshold, the model adaptive optimization mechanism will be triggered;

[0053] The optimization process adjusts the model parameters based on error integration to reduce the difference between the predicted value and the actual value. The verified prediction curve will be packaged into the final version of the vibration response prediction curve data package.

[0054] Optionally, the S5 includes:

[0055] S51: Analyze the vibration response prediction curve output by S44, extract the peak vibration amplitude sequence in the curve, and extract vibration characteristics for different transportation conditions:

[0056] In the land working condition, the maximum vibration amplitude in the 0–20 Hz frequency band is extracted;

[0057] In the track working condition, the maximum acceleration peak value of each impact point is extracted;

[0058] During the mode switching process, the maximum amplitude of the transient envelope is extracted to characterize the degree of sudden vibration caused by the working condition switching;

[0059] S52: establishing a vibration amplitude-pressure conversion model to convert the peak vibration amplitude extracted from the vibration response prediction curve into an equivalent pressure value;

[0060] The vibration amplitude-pressure conversion model establishes different mapping relationships for land mode, track mode, and mode switching. Combined with tomato quality parameters during actual transportation, it quantitatively calculates the equivalent mechanical pressure under different vibration conditions, providing a unified metric for subsequent comparison with damage thresholds.

[0061] S53: The converted equivalent pressure value is compared with the damage threshold output by S33 in real time to calculate the safety margin under the current transportation status. This safety margin serves as a core indicator to measure whether the vibration experienced by tomatoes during transportation exceeds their physical tolerance limit. It can dynamically reflect the potential risk level of the current driving speed to the fruit safety.

[0062] Optionally, the S5 further includes:

[0063] S54: A speed adjustment decision tree is established to determine whether the transport vehicle needs to adjust its speed based on the calculated safety margin value. The decision tree has three conditional branches:

[0064] If the safety margin is within a reasonable range, maintain the current speed;

[0065] If the safety margin approaches the threshold, a proportional speed reduction strategy is implemented;

[0066] If the safety margin is negative, it is considered a high-risk state and emergency braking must be performed immediately to reduce the speed to a safe level to prevent damage to the tomato due to excessive vibration.

[0067] S55: Generate a maximum permissible speed value for each road section according to the output result of the speed adjustment decision tree.

[0068] If the road section is determined to be safe, the maximum allowed speed value is taken as the current road speed limit;

[0069] In sections with potential risks, the decision speed is compared with the road speed limit, and the smaller value is selected as the maximum allowable speed for the section to ensure that the transportation process operates within the safety constraint range;

[0070] S56: Bind the maximum allowable speed value corresponding to each road section to the spatial coordinates of the section to construct a speed decision matrix. This matrix is ​​organized according to geographic coordinates or time series and is used to provide a clear and structured speed reference standard for the segmented speed control in step S6, achieving efficient docking and execution of dynamic control logic.

[0071] Optionally, the S6 includes:

[0072] S61: Analyze the speed decision matrix generated by S56 and extract the maximum allowable speed values ​​and the corresponding road segment boundary coordinates of all road segments within 3 kilometers ahead of the current driving direction. This process is used to construct a local driving planning window and establish a speed control range in advance for the target road segment to be entered, ensuring that the control instructions are predictable and continuous.

[0073] S62: Calculate a pre-deceleration start point based on the distance difference between the vehicle's real-time positioning coordinates and the boundary coordinates of each road segment in the speed decision matrix. This start point comprehensively considers factors such as the current vehicle speed, the target speed, and the vehicle's maximum deceleration rate, and adds a fixed safety margin to determine when to start speed control to ensure smooth and safe speed transitions.

[0074] S63: When the distance between the vehicle and the target road section boundary is less than or equal to the pre-deceleration starting point, a step-by-step speed change control command is generated. This command includes key fields such as command type, starting speed, target speed, acceleration, and effective distance. It is used to instruct the vehicle control system to complete a smooth transition from the current speed to the maximum allowable speed value within the specified distance, ensuring that the control process meets both route safety and fruit stability.

[0075] S64: The step-shift control command is sent to the electronically controlled throttle and brake pressure units via the CAN bus, enabling coordinated speed control. The throttle opening is adjusted incrementally based on the system control model, exhibiting a nonlinear attenuation trend over time. Brake pressure is linearly adjusted proportionally based on the difference between the current vehicle speed and the target speed, forming a dynamic coordination mechanism between the throttle and brake systems for highly precise speed control.

[0076] S65: continuously monitoring in real time the deviation between the actual speed of the vehicle and the maximum permissible speed value during the execution process;

[0077] If the absolute value of the deviation is detected to be greater than 5 kilometers per hour and lasts for more than 2 seconds, it will be determined as a deviation exceeding the limit state and the instruction regeneration process will be immediately triggered. This process includes re-calling S55 to update the maximum allowable speed value and re-executing the control process from S61 to S64 to ensure that the control system has adaptive and error correction capabilities;

[0078] S66: When the vehicle passes through the boundary of the target road section, the execution status of this segmented control is compiled into an instruction execution report and reported to the monitoring center in real time through the communication module. The report includes the vehicle's actual speed curve, the target speed curve, and the fit index between the two. It is used to quantitatively evaluate the effect of this control and provide data support for subsequent operation control optimization and full traceability management.

[0079] Beneficial effects of the present invention:

[0080] This invention uses multi-source sensing units to acquire physical characteristic data such as road images, vibration waveforms, and altitude. Using methods such as image recognition, frequency domain analysis, and differential calculation, it constructs a time-stamped, real-time road condition information package covering three parameters: road type, road surface roughness, and altitude gradient. Based on this, the system can rapidly identify the driving mode during transportation, including land mode, track mode, and their switching states, and extract bump parameters such as vertical vibration frequency, track impact intensity, and attitude adjustment time. This enables precise classification and structured representation of dynamic operating conditions under complex transportation routes, providing a reliable basis for subsequent speed control.

[0081] This invention uses multispectral imaging and thermal imaging to identify the maturity level of tomatoes. Combining a maturity-compressive strength correspondence table with a stacking layer correction factor, it dynamically calculates the damage threshold for each batch of tomatoes in a container. Furthermore, a modified Duffing equation, a Fourier series model, and a transient response function are constructed for different driving modes to establish a mapping relationship between vibration parameters and equivalent pressure. This dynamic mapping model not only outputs a high-resolution vibration response prediction curve but also compares the converted equivalent pressure with the damage threshold in real time, quantifying the safety margin at each instant during transportation and effectively improving the accuracy of fruit damage resistance assessment.

[0082] The present invention establishes a speed adjustment decision tree based on the difference between the damage threshold and the equivalent pressure, and on this basis generates the maximum allowable speed value for each road section; at the same time, it generates pre-deceleration points through the real-time positioning information of the vehicle combined with the predicted path, and dynamically generates segmented speed change control instructions to control the throttle opening and brake pressure to perform gradual speed adjustment. The system supports triggering automatic instruction reconstruction under the condition of excessive speed deviation, and has adaptive deviation correction and closed-loop adjustment capabilities. In addition, through the instruction execution report reporting mechanism, the control process can be tracked, recorded and evaluated for feedback, providing technical support for the full-link safety and controllability of the intelligent transportation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0084] Figure 1 Schematic diagram of a method flow in an embodiment of the present invention;

[0085] Figure 2 This is a schematic diagram of the S5 process of an embodiment of the present invention. DETAILED DESCRIPTION

[0086] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0087] like Figure 1-2 As shown, a method for adaptively adjusting the tomato transportation speed based on path road condition feedback includes the following steps:

[0088] S1: Obtain real-time road condition information of the target transport route, including road type, road surface flatness, and altitude gradient;

[0089] S2: Identify the current driving mode based on real-time road condition information. Driving modes include track mode and land mode, and calculate corresponding bump parameters, including vertical vibration frequency, lateral swing amplitude and duration in land mode, track joint impact strength and upper and lower rail transition section inclination angle in track mode, as well as maximum impact acceleration and attitude adjustment time when switching modes.

[0090] S3: Based on the tomato maturity detection data and the preset maturity-compressive strength correspondence table, the damage threshold of the tomatoes in the current cargo box is calculated;

[0091] S4: Establish a dynamic mapping model between cargo box vibration amplitude and bump parameters, distinguish between track and land conditions and generate vibration response prediction curves;

[0092] S5: Determine the maximum permissible speed value for each road section based on the difference between the vibration response prediction curve and the damage threshold;

[0093] S6: Generate a segmented speed change control command based on the maximum allowable speed value and perform a pre-deceleration operation before entering the target section.

[0094] S1 includes:

[0095] S11: Synchronously collect physical feature data of the target transportation path through a multi-source sensing unit. The physical feature data includes a road image sequence captured by a machine vision module, a road surface vibration waveform measured by a three-axis accelerometer, and an altitude point set recorded by an RTK-GPS module.

[0096] S12: Perform convolutional neural network classification processing on the road image sequence and output structured road type parameters. The road type parameters include classification codes and corresponding confidence levels for asphalt pavement, cement pavement, and gravel pavement. The classification function is expressed as follows:

[0097] ;

[0098] in, represents the road image frame collected at time t, It is the feature extraction and classification function of the trained convolutional neural network. Respectively represent the classification probabilities of asphalt, cement, and gravel roads, satisfying ;

[0099] S13: Perform frequency domain energy analysis on the road vibration waveform, extract the vibration root mean square value in the 5-20 Hz frequency band, and convert it into a standard road surface roughness index through a calibration curve. The road surface roughness index unit is mm / m. The vibration root mean square value is calculated as follows:

[0100] ;

[0101] ;

[0102] in, is the acceleration amplitude at frequency fHz, N is the number of sampling points in the frequency band, is the RMS value of the frequency band, is the converted road surface roughness index, ,b is the linear mapping coefficient obtained through calibration experiments;

[0103] S14: Perform sliding window difference calculation on the elevation point set to generate a continuous elevation change rate curve. Extract the road section data with an absolute value of the elevation change rate greater than 3% in the curve as the effective elevation change gradient parameter. The calculation formula is as follows:

[0104] ;

[0105] in, is the altitude of the i-th point, is the distance position of the i-th point in the path, w is the step size of the sliding window, is the rate of change of altitude per unit distance within the window;

[0106] S15: The road type parameter, road surface roughness index, and altitude change gradient parameter are integrated according to a unified spatiotemporal benchmark to generate a real-time road condition information data packet with a timestamp, wherein the road type, road surface roughness, and altitude change gradient parameters are updated at a frequency of 10 Hz per second.

[0107] S2 includes:

[0108] S21: Execute mode decision logic according to the road type parameter in the real-time road condition information generated in S15: trigger the track mode when the road type is a railway track, otherwise trigger the land mode;

[0109] S22: Perform bump parameter calculation in land mode:

[0110] Based on the S15 road surface roughness index, the vertical vibration frequency is calculated using the following formula:

[0111] ;

[0112] in, is the vertical vibration frequency, IRI is the road surface roughness index,

[0113] Based on the altitude gradient parameter of S15, the lateral swing amplitude is calculated using the following formula:

[0114] ;

[0115] in, is the lateral swing amplitude, G is the slope percentage (i.e., the gradient of altitude change), and L is the wheelbase of the truck;

[0116] Extract the length of the road section where the absolute value of the slope is continuously ≥3%, and deduce the duration of the slope based on the current vehicle speed:

[0117] ,in, The duration of the turbulence, is the length of the continuous road section with an absolute slope of ≥3%, and v is the current vehicle speed;

[0118] S23: Perform bump parameter calculation in orbit mode:

[0119] According to the track joint spacing m and current vehicle speed v, calculate the wheel-rail impact period :

[0120] ;

[0121] At the same time, the acceleration peak value within the cycle is extracted , as the impact strength of the track joint:

[0122] ;

[0123] in, is the impact strength of the track joint, is the longitudinal acceleration data recorded during the impact cycle.

[0124] Based on the elevation difference between adjacent track sections , calculate the inclination angle of the transition section between the upper and lower rails by the following formula:

[0125] ;

[0126] in, is the inclination angle of the transition section between the upper and lower rails, is the actual horizontal length between the two track segments, is the altitude difference between track sections;

[0127] S24: When the mode is switched (land-to-track or track-to-land):

[0128] Capture the maximum peak value of the acceleration sensor at the switching moment as the maximum impact acceleration:

[0129] ;

[0130] in, is the maximum impact acceleration, Mode switching time point, It is the monitoring window time before and after;

[0131] Record the time from when the switching command is issued to when the vehicle body posture angle changes and stabilizes within ±1°, and record it as the posture adjustment time. ;

[0132] S25: Encapsulates the parameters calculated by S22 / S23 / S24 into a structured turbulence parameter data packet, including:

[0133] Land mode parameters: vertical vibration frequency , lateral swing amplitude and duration of action ;

[0134] Track mode parameters: Track joint impact strength Inclination angle of transition section between upper and lower rails ;

[0135] Mode switching parameters: Maximum impact acceleration Time-consuming posture adjustment .

[0136] S3 includes:

[0137] S31: The multispectral imaging unit on the top of the cargo box scans the surface of the tomatoes, collecting the dual-band reflectance ratio of 520nm and 680nm wavelengths. This is combined with the infrared thermal image to generate tomato maturity detection data. This data includes the maturity level (1-5) of each tomato and the spatial distribution matrix.

[0138] Maturity level is determined based on multispectral reflectance ratio Temperature distribution with thermal map The classification model constructed by the combined threshold rule generates a two-dimensional space matrix: ,in and The unit reflectances are 520nm and 680nm respectively;

[0139] The maturity level is derived from multi-threshold segmented mapping and automatically assigned a level of 1-5.

[0140] S32: Calling the maturity-compressive strength correspondence table pre-stored in the control system, which is a two-dimensional array structure:

[0141] Maturity Level 1 2 3 4 5 Compressive strength (MPa) 0.85 0.72 0.58 0.41 0.33

[0142] Use the tomato maturity level output by S31 as the row index and perform a batch table lookup to obtain the corresponding compressive strength value. ,in ,N is the total number of tomatoes in the cargo box;

[0143] S33: Based on the spatial distribution matrix of tomatoes in the cargo box, take the minimum compressive strength value of all tomatoes as the baseline compressive strength , and calculate the damage threshold of tomatoes in the current container according to the following formula :

[0144] ;

[0145] in, is the damage threshold of tomatoes in the box, in MPa, is the minimum compressive strength of all tomatoes, H is the number of stacked layers of tomatoes in the cargo box, which is a non-negative integer, 1.25 is the safety correction factor, and 0.05 is the stacking layer influence coefficient.

[0146] S34: Synchronously encapsulate the damage threshold and the bumpy parameters output by S25 into a cargo state feature vector, which is used to represent the transportation tolerance state at a specific time point and is called by step S4 to perform dynamic mapping modeling.

[0147] S4 includes:

[0148] S41: Parse the bump parameters output by S25 and separate the input data set according to the mode identifier in the parameter:

[0149] Land mode data set: including vertical vibration frequency, lateral oscillation amplitude and duration;

[0150] Track pattern dataset: contains track joint impact strength and the inclination angle of the transition section between the upper and lower rails;

[0151] Mode switching dataset: contains the maximum impact acceleration and attitude adjustment time;

[0152] S42: Construct a dynamic mapping model with decoupled working conditions to predict the vibration response characteristics caused by different working conditions in the cargo box:

[0153] Land model: The Duffing equation is used for modeling, which is as follows:

[0154] ;

[0155] in, is the vibration response intensity in land mode, is the vertical vibration frequency, is the horizontal swing amplitude, For the duration of action, It is an empirical coefficient calibrated by bench test.

[0156] Orbital mode model: constructed based on Fourier series expansion, the form is:

[0157] ;

[0158] in, is the vibration response intensity in track mode, is the impact strength of the track joint, is the inclination angle of the transition section between the upper and lower rails, are the amplitude coefficient and exponential decay factor of the Fourier series, which are preset by experiments.

[0159] Mode switching model: The transient impulse response function is used for modeling, which is in the form of:

[0160] ;

[0161] in, is the transient vibration value at t seconds after mode switching, is the maximum impact acceleration, is the time taken for posture adjustment, t is the time variable, ranging from [0,3] seconds;

[0162] S43: Input the data sets separated in S41 into the corresponding models to generate the vibration response prediction curve under the working condition:

[0163] Land conditions: Output vibration amplitude-time curve in the 0–20 Hz frequency range;

[0164] Track working condition: Output the vibration peak sequence at each track joint impact point

[0165] Mode switching condition: Output the transient vibration envelope within 3 seconds after switching, reflecting the vibration changes of the vehicle body during the transition period;

[0166] Each prediction curve is high-frequency time domain data with a sampling rate of 1000 points / second, ensuring the capture of dynamic changes at the microsecond level;

[0167] S44: Execute the model validation and optimization process: First, the predicted curve is compared point by point with the measured data collected by the on-board accelerometer. When the relative error within a certain period exceeds 15%, the adaptive model tuning is triggered;

[0168] The tuning process adjusts the model response by error integration. The updated dynamic mapping model is expressed as follows:

[0169] Dynamic Mapping Model Dynamic Mapping Model Error integral);

[0170] The error integral is the accumulation of the absolute difference between the measured curve and the predicted curve in the time domain. is the built-in response weight adjustment function;

[0171] The final output is a verified vibration response prediction curve data package, which serves as the input reference for the speed adjustment module.

[0172] S5 includes:

[0173] S51: Analyze the vibration response prediction curve output by S44 and extract the peak vibration amplitude sequence in the curve, where:

[0174] The maximum vibration amplitude of the 0-20 Hz frequency band is extracted under land conditions and recorded as ;

[0175] The maximum acceleration peak value of the impact point extracted from the track working condition is recorded as ;

[0176] The maximum amplitude of the transient envelope extracted by mode switching is recorded as ;

[0177] S52: Establish a vibration amplitude-pressure conversion model to convert the peak vibration amplitude into an equivalent pressure value. The conversion rules are as follows:

[0178] ;

[0179] in, is the equivalent pressure value, is the peak vibration amplitude in the land mode, is the maximum acceleration peak at the track joint, is the maximum transient vibration amplitude during the mode switching process, is the average mass of a single tomato;

[0180] S53: The converted equivalent pressure value is compared with the damage threshold output by S33 in real time to calculate the safety margin under the current transportation state. The calculation method is as follows:

[0181] ;

[0182] in, is the safety margin, is the damage threshold of tomatoes in the box calculated in S33, is the maximum equivalent pressure value under the current road section working conditions;

[0183] S54: Establish a speed adjustment decision tree based on the current safety margin value Determine the speed control strategy required for transport vehicles. The decision logic is as follows:

[0184] like MPa, the current speed can be maintained unchanged;

[0185] like MPa, then reduce the current vehicle speed to 0.8 times;

[0186] like MPa, it is judged as high risk and emergency braking is required to reduce the speed to a safe speed.

[0187] S55: Based on the output of the decision tree, the maximum permissible speed value for each road section is generated. Specifically:

[0188] If the road section is determined to be a safe road section, the maximum allowed speed value Equal to the road speed limit of the road section;

[0189] If a road section is judged to be risky, the maximum permissible speed value The smaller value between the output speed value and the road speed limit value is used for decision making;

[0190] S56: Bind the maximum allowable speed value of each road section with the corresponding road section coordinates to construct a speed decision matrix. The matrix is ​​arranged in chronological order or geographical coordinate order to provide a complete speed benchmark for performing segmented speed change control in step S6.

[0191] S6 includes:

[0192] S61: Analyze the speed decision matrix generated by S56, extract the maximum allowable speed values ​​of all road sections within 3 kilometers ahead of the current driving direction and the corresponding road section boundary coordinates, and use them to construct a local driving planning window to prepare for the preset speed adjustment of the target road section to be entered;

[0193] S62: Calculate the pre-deceleration starting point based on the distance difference between the vehicle's real-time positioning coordinates and the road segment boundary coordinates. The calculation logic includes converting the kinetic energy difference between the current vehicle speed and the target speed into the required deceleration distance, and adding a fixed safety margin distance. The formula is:

[0194] ,in, is the distance between the pre-deceleration starting point and the road section boundary, is the current vehicle speed, is the target maximum allowable speed value, The maximum deceleration rate, the default value is -3m / s2, This is the redundant distance compensation value, fixed at 50 m. If the system meter value is less than 200 m, 200 m will be used as the minimum pre-deceleration control trigger threshold;

[0195] S63: When the distance between the vehicle and the road segment boundary is less than or equal to When the vehicle is in the state of acceleration, a segmented speed change control instruction is generated. The control instruction includes key parameters such as starting speed, target speed, acceleration value and effective distance, which makes it easier for the vehicle control system to smoothly adjust the speed according to the set curve. Determined by the following mathematical formula:

[0196] ; is the remaining distance from the current position to the boundary of the road segment;

[0197] S64: The above-mentioned step-speed shift control command is sent to the electronically controlled throttle and brake pressure unit via the CAN bus to perform the actual speed control action. The throttle opening adopts exponential decay control in the form of:

[0198] ;

[0199] in, is the current throttle opening, is the initial throttle opening, t is the execution time, is the throttle response time constant.

[0200] The brake pressure is adjusted in a proportional control manner as follows:

[0201] ;in, is the current brake pressure, is the braking proportional gain coefficient, Indicates the current vehicle speed and target vehicle speed;

[0202] S65: The system monitors the vehicle's current speed in real time Maximum permissible speed The deviation between , , when the following two conditions are met:

[0203] km / h and duration Second;

[0204] The system triggers the instruction regeneration process, which includes the following steps:

[0205] Call back S55 and recalculate the safety margin based on the latest equivalent pressure value and damage threshold;

[0206] Update the maximum allowed speed value for each road segment ;

[0207] Re-execute the speed control logic from S61 to S64 to match the updated speed limit and driving conditions;

[0208] This process ensures that control instructions are adaptive and can perform rapid dynamic corrections when the vehicle's actual operating state deviates from the predicted model.

[0209] S66: When a vehicle passes through the target road section boundary, the system compiles the results of this segmented control into a command execution report and transmits it to the monitoring center via the communication module. This report includes the vehicle's actual speed curve, the preset target speed curve, and the fit between the two. This report serves as a quantitative assessment of the quality of control throughout the transportation process and is used for subsequent operation control optimization and traceability audits.

[0210] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0211] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for adaptively adjusting tomato transportation speed based on route and road condition feedback, characterized in that: The following steps are involved: S1: Acquire real-time road condition information of the target transport route, wherein the road condition information includes road type, road surface flatness, and altitude gradient; S2: Identify the current driving mode based on real-time road condition information. Driving modes include track mode and land mode, and calculate corresponding bump parameters, including vertical vibration frequency, lateral swing amplitude and duration in land mode, track joint impact strength and upper and lower rail transition section inclination angle in track mode, as well as maximum impact acceleration and attitude adjustment time when switching modes. S3: Based on the tomato maturity detection data and the preset maturity-compressive strength correspondence table, the damage threshold of the tomatoes in the current cargo box is calculated; S4: Establish a dynamic mapping model between cargo box vibration amplitude and bump parameters, distinguish between track and land conditions and generate vibration response prediction curves; S5: Determine the maximum permissible speed value for each road section based on the difference between the vibration response prediction curve and the damage threshold; S6: Generate a segmented speed change control command based on the maximum allowable speed value and perform a pre-deceleration operation before entering the target section.

2. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 1, characterized in that: Said S1 comprises: S11: Synchronously collecting physical feature data of the target transportation path through a multi-source sensing unit, wherein the physical feature data includes a road image sequence captured by a machine vision module, a road surface vibration waveform measured by a three-axis accelerometer, and an altitude point set recorded by an RTK-GPS module; S12: Performing convolutional neural network classification processing on the road image sequence to output structured road type parameters, wherein the road type parameters include classification codes and corresponding confidence levels of asphalt pavement, cement pavement, and gravel pavement; S13: Performing frequency domain energy analysis on the road surface vibration waveform, extracting the vibration root mean square value in the 5-20 Hz frequency band, and converting it into a standard road surface roughness index through a calibration curve; S14: performing a sliding window difference calculation on the elevation point set to generate a continuous elevation change rate curve, and extracting road section data with an absolute value of a change rate greater than 3% in the curve as an effective elevation change gradient parameter; S15: The road type parameter, road surface roughness index, and altitude gradient parameter are integrated according to a unified time-space reference to generate a real-time road condition information data packet with a timestamp.

3. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 2, characterized in that: The S2 includes: S21: Execute mode decision logic based on the road type parameter in the real-time road condition information generated in S15: if the road type is a railway track, trigger the track mode; otherwise, trigger the land mode; S22: Calculate the bump parameters in the land mode and calculate the corresponding vertical vibration frequency based on the road surface roughness index provided by S15; Using the altitude gradient parameter in S15 and the vehicle wheelbase, the lateral swing amplitude of the vehicle under the slope is calculated; Identify road sections with a slope of 3% or more, and deduce the duration of the vehicle's action on that type of slope based on the current vehicle speed. S23: performing bump parameter calculation in track mode: calculating the wheel-rail impact cycle based on the fixed track joint spacing and the current vehicle speed, and extracting the acceleration peak within the cycle range as the track joint impact intensity; At the same time, the elevation difference between adjacent track segments is extracted, and combined with the track segment length, the inclination angle of the upper and lower track transition section is calculated to evaluate the posture change trend of the vehicle in track mode.

4. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 3, characterized in that: Said S2 further comprises: S24: If a mode switch occurs during transportation, the maximum acceleration value at the switching moment is captured by the acceleration sensor and recorded as the maximum impact acceleration; Secondly, the time from when the switching command is issued to when the vehicle's attitude angle stabilizes within a range of plus or minus 1 degree is tracked. This time is the attitude adjustment time and is used to evaluate the transition impact caused by mode switching. S25: The turbulence parameters calculated in S22, S23, and S24 are uniformly encapsulated into a structured turbulence parameter data package. The data package includes: vertical vibration frequency, lateral swing amplitude, and duration of action in the land mode parameters; track joint impact strength and inclination angle of the upper and lower rail transition sections in the track mode parameters; and maximum impact acceleration and attitude adjustment time in the mode switching parameters.

5. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 4, characterized in that: The S3 includes: S31: The multispectral imaging unit on top of the cargo box scans the surface of the tomatoes, collecting the dual-band reflectance ratio of 520nm and 680nm wavelengths. This is combined with infrared thermal image information to identify the maturity of each tomato. Generate tomato maturity detection data based on preset image recognition logic. The tomato maturity detection data includes the maturity level of each tomato, which is divided into levels 1 to 5. At the same time, a spatial distribution matrix is ​​generated that reflects the spatial location of tomatoes at each level. S32: calling a pre-stored maturity-compressive strength correspondence table, where the maturity-compressive strength correspondence table presents compressive strength values ​​corresponding to different maturity levels in a two-dimensional array format; According to the tomato maturity level output in S31, the maturity-compressive strength correspondence table is called one by one, and the maturity level of each tomato is used as the table index to batch match and extract the corresponding compressive strength values ​​of all tomatoes; S33: Combining all the compressive strength values ​​in the tomato spatial distribution matrix, extracting the minimum compressive strength value as the baseline compressive strength value, and correcting the baseline compressive strength value based on the preset damage threshold calculation rules and the number of stacked layers of tomatoes in the cargo box to obtain the damage threshold of the tomatoes in the current cargo box; S34: The calculated damage threshold and the bump parameter data output in S25 are synchronously encapsulated into a cargo state feature vector.

6. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 5, characterized in that: The S4 includes: S41: parsing the bump parameters output by S25, and classifying the parameter contents into three independent data sets according to the mode identification fields therein; The land mode data set includes three indicators: vertical vibration frequency, lateral oscillation amplitude and duration of action; The track pattern dataset includes the impact strength of track joints and the inclination angle of the transition section between the upper and lower rails; The mode switching data set includes the maximum impact acceleration and attitude adjustment time; S42: Based on the data grouping, a dynamic mapping model with decoupled working conditions is constructed. The dynamic mapping model consists of three sub-models, corresponding to the land mode, track mode, and mode switching scenario respectively; The land mode model predicts the vibration response of the vehicle in this state through vertical vibration frequency, lateral swing amplitude and duration; The track mode model describes the response to impact accumulation through the impact strength of the track joint and the inclination angle of the transition section between the upper and lower rails; The mode switching model describes the transient vibration changes generated when the vehicle transitions from one mode to another based on the maximum impact acceleration and the posture adjustment time. S43: Inputting various data sets separated in S41 into corresponding dynamic mapping models to generate vibration response prediction curves under land conditions, track conditions, and mode switching conditions, respectively; The output of the land working condition is the vibration amplitude-time curve covering the frequency range of 0 to 20 Hz; Track working condition outputs the vibration peak sequence of the track joint impact point; Transient vibration envelope within 3 seconds after the output is switched in the mode switching condition; S44: performing a model validation and optimization process on the generated vibration response prediction curve, comparing the prediction curve with the measured data obtained by the vehicle accelerometer during the actual transportation process point by point. If the relative error within a certain period exceeds a set threshold, the model adaptive optimization mechanism will be triggered; The optimization process adjusts the model parameters based on error integration to reduce the difference between the predicted value and the actual value. The verified prediction curve will be packaged into the final version of the vibration response prediction curve data package.

7. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 6, characterized in that: The S5 includes: S51: Analyze the vibration response prediction curve output by S44, extract the peak vibration amplitude sequence in the curve, and extract vibration characteristics for different transportation conditions: In the land working condition, the maximum vibration amplitude in the 0–20 Hz frequency band is extracted; In the track working condition, the maximum acceleration peak value of each impact point is extracted; During the mode switching process, the maximum amplitude of the transient envelope is extracted; S52: establishing a vibration amplitude-pressure conversion model to convert the peak vibration amplitude extracted from the vibration response prediction curve into an equivalent pressure value; The vibration amplitude-pressure conversion model sets different mapping relationships for land mode, track mode, and mode switching. Combined with the actual tomato quality parameters during transportation, it can quantitatively calculate the equivalent mechanical pressure under different vibration conditions. S53: The converted equivalent pressure value is compared with the damage threshold output by S33 in real time to calculate the safety margin under the current transportation state.

8. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 7, characterized in that: The S5 further includes: S54: establishing a speed adjustment decision tree, and determining whether the transport vehicle needs to adjust its speed based on the calculated safety margin value; S55: generating a maximum permissible speed value for each road section according to the output result of the speed adjustment decision tree; S56: Bind the maximum allowable speed value corresponding to each road section to the spatial coordinates of the road section to construct a speed decision matrix.

9. The method for adaptively adjusting tomato transportation speed based on route and road condition feedback according to claim 8, characterized in that: The S6 includes: S61: Analyze the speed decision matrix generated in S56, and extract the maximum allowable speed values ​​of all road sections within 3 kilometers ahead in the current driving direction and the corresponding road section boundary coordinates; S62: Calculating a pre-deceleration starting point based on the distance difference between the real-time positioning coordinates of the vehicle and the boundary coordinates of each road section in the speed decision matrix; S63: When the distance between the vehicle and the target road section boundary is less than or equal to the pre-deceleration starting point, a segmented speed change control instruction is generated; S64: Sends the segmented speed control command to the electronically controlled throttle and brake pressure unit via the CAN bus to implement the actual coordinated speed control execution; S65: continuously monitoring in real time the deviation between the actual speed of the vehicle and the maximum permissible speed value during the execution process; If the absolute value of the deviation is detected to be greater than 5 kilometers per hour and lasts for more than 2 seconds, it will be determined as a deviation exceeding the limit state and the instruction regeneration process will be immediately triggered; S66: When the vehicle passes through the boundary of the target road section, the execution status of this segment control is compiled into an instruction execution report and reported to the monitoring center in real time through the communication module.

Images