Unmanned logistics vehicles, load health management methods, and cloud-edge collaboration platforms
By collecting and evaluating the operational data of the unmanned logistics vehicle load in real time through the chassis domain controller, and using health assessment models and remaining life prediction models, the problem of early anomaly identification of the on-board load of unmanned logistics vehicles is solved, achieving efficient and accurate load health management and improving operational efficiency and safety.
Patent Information
- Application Number
- CN202611349097.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-01
- Publication Date
- 2026-11-10
AI Technical Summary
Existing technologies struggle to identify early anomalies in the load of unmanned logistics vehicles in a timely and accurate manner, resulting in low fault location efficiency, inaccurate maintenance decisions, high maintenance costs, and a high risk of unplanned downtime.
The chassis domain controller of the unmanned logistics vehicle collects raw operating data of multiple target loads in real time, uses a health assessment model to evaluate the health of the loads, and predicts the types of load failures and spare parts required for maintenance based on the remaining life prediction model, generating maintenance memos to achieve non-intrusive load health management.
It improves the adaptability and accuracy of load health assessment, automatically identifies fault types, accurately determines the spare parts required for maintenance, forms a closed loop for load health management, maximizes the operational efficiency of unmanned logistics vehicles, and ensures operational safety.
Smart Images

Figure CN123067463A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of vehicle load health assessment technology, and in particular to an unmanned logistics vehicle, a load health management method, and a cloud-edge collaborative platform. Background Technology
[0002] With the application of autonomous driving technology in scenarios such as park transportation, last-mile delivery, and cold chain logistics, autonomous logistics vehicles are typically equipped with various onboard loads, including lights, horns, fans, water pumps, solenoid valves, electronic locks, and refrigeration units. The operational status of these onboard loads directly affects vehicle driving safety, operational performance, and cargo transportation. Since autonomous logistics vehicles lack on-site driver inspections during operation, timely detection and effective maintenance of abnormal onboard load conditions have become crucial issues in the operation and management of autonomous logistics vehicles.
[0003] Existing vehicle load maintenance methods mainly include post-fault repair and periodic maintenance / replacement. Post-fault repair typically involves the remote monitoring system detecting the fault based on abnormal task execution after the vehicle load malfunctions, followed by dispatching maintenance personnel to inspect and replace the faulty component. Some vehicle controllers can generate and report corresponding fault codes when abnormalities such as open circuits or short circuits occur in the load circuit, allowing maintenance personnel to read and handle them upon the vehicle's return to the repair shop. Periodic maintenance / replacement primarily targets critical or time-limited loads such as water pumps, solenoid valves, and electronic locks, performing mandatory inspections or replacements according to preset mileage or time cycles.
[0004] However, post-failure repair methods typically only address issues after the onboard load has exhibited obvious abnormalities or functional failures, making it difficult to promptly identify early anomalies such as gradual performance degradation, intermittent poor contact, or occasional overcurrent. For autonomous logistics vehicles, sudden onboard load failures can disrupt transportation or operations, requiring emergency dispatch for rescue. For example, a malfunctioning electronic lock may prevent the container from opening properly, and a refrigeration unit failure may affect the storage and transportation environment of goods, leading to spoilage. After a failure occurs, maintenance personnel must systematically check the load, power supply lines, controllers, and fuses, resulting in lengthy fault location and repair times and impacting vehicle operational efficiency.
[0005] Regular maintenance and replacement methods typically use uniform mileage or time cycles as the basis for maintenance, failing to fully consider the differences in actual operating time, start-stop frequency, load current, operating environment, and performance degradation of different vehicle loads. Therefore, some vehicle loads that are still in normal condition may be replaced prematurely, resulting in a waste of maintenance resources; some vehicle loads may also fail before the scheduled maintenance cycle arrives, making it difficult to effectively prevent unplanned downtime through regular maintenance.
[0006] Furthermore, existing maintenance methods typically lack continuous data collection and trend analysis of onboard load operations, making it difficult to assess the health of the onboard load or predict its performance trends based on its actual operating status. Consequently, maintenance personnel struggle to develop targeted maintenance plans in advance. Especially when autonomous logistics vehicles operate far from maintenance bases, load failures can lead to problems such as longer rescue distances, longer repair response times, and a lack of targeted spare parts availability.
[0007] Therefore, existing technologies struggle to identify early anomalies in the onboard load of autonomous logistics vehicles and assess their health status in a timely and accurate manner, resulting in low fault location efficiency, inaccurate maintenance decisions, high maintenance costs, and a higher risk of unplanned downtime. Summary of the Invention
[0008] In view of the shortcomings of the prior art described above, the purpose of this application is to provide an unmanned logistics vehicle, a load health management method, and a cloud-edge collaborative platform to solve the technical problems of low fault location efficiency, inaccurate maintenance decisions, high maintenance costs, and high risk of unplanned downtime caused by the inability of the prior art to identify early anomalies of the on-board load of the unmanned logistics vehicle in a timely and accurate manner and to assess its health status.
[0009] To achieve the above and other related objectives, a first aspect of this application provides a method for managing the load health of an unmanned logistics vehicle. This method is applied to the chassis domain controller of the unmanned logistics vehicle, which is connected to the autonomous driving domain controller and the operation and maintenance platform of the vehicle. The method includes: real-time collection of raw operating data of multiple target loads in the unmanned logistics vehicle, and feature extraction of each raw operating data to generate health feature data for each target load; assessment of the health of each target load based on a pre-built health assessment model for each target load, according to the health feature data; prediction of the remaining service life of each target load based on a pre-built remaining service life prediction model for each target load, identification of the fault type of each target load, determination of the spare parts required for maintenance of each target load, generation of a maintenance memo for each target load, and reporting to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load, and the autonomous driving domain controller can adjust the vehicle task strategy of the unmanned logistics vehicle.
[0010] In some embodiments of the first aspect of this application, the original operating data includes at least one or more combinations of: switching event monitoring data of the corresponding target load within the current time window, continuous operation monitoring data, and environmental compensation data; wherein, the switching event monitoring data includes at least one or more combinations of: the number of switching on events of the corresponding target load within the current time window, the peak value and duration of the surge current generated by each switching on event, the number of switching off events, and the back EMF drop waveform of each switching off event; the continuous operation monitoring data includes at least: the energizing time of the corresponding target load within the current time window, and the steady-state current value and on-state voltage drop data at multiple sampling times; the environmental compensation data includes at least: ambient temperature and battery voltage; the health characteristic data includes at least: surge current health characteristic data, steady-state current health characteristic data, and on-state voltage drop data of the corresponding target load within the current time window. The surge current health characteristic data includes at least one or more of the following: surge current health characteristic data, back EMF decay health characteristic data, cumulative switching count, and cumulative running time; wherein, the surge current health characteristic data includes at least one or more of the following: surge current impact coefficient and maximum surge current impact coefficient for each switching on event of the target load within the current time window; the steady-state current health characteristic data includes at least one or more of the following: steady-state current relative deviation at multiple sampling times of the target load within the current time window, and steady-state current average value, steady-state current peak-to-peak value, steady-state current ripple rate, and steady-state current average relative deviation within the current time window; the on-state voltage drop health characteristic data includes at least one of the following: on-state voltage drop trend of the target load within the current time window; the back EMF decay health characteristic data includes at least one of the following: back EMF decay time constant and maximum back EMF decay time constant for each switching off event of the target load within the current time window.
[0011] In some embodiments of the first aspect of this application, the various health assessment models employ the same mathematical model framework, which can be expressed as: ;in, The health status of the target load is expressed as a percentage. The first health sub-item of the target load is determined based on the ratio between the average relative deviation of the steady-state current of the target load within the current time window and the allowable current deviation of the target load. The second health sub-item of the target load is determined based on the ratio between the maximum surge current impact factor of the target load within the current time window and the allowable surge current impact factor of the target load. The third health sub-item of the target load is determined based on the ratio between the peak-to-peak steady-state current of the target load within the current time window and the allowable current fluctuation range of the target load. The fourth health sub-item of the target load is determined based on the on-state voltage drop trend of the target load within the current time window; The fifth health sub-item of the target load is determined based on the changing trend of the maximum back EMF decay time constant of the target load within the current time window; The sixth health sub-item of the target load is determined as a percentage between the cumulative number of switching operations or the cumulative power-on time and the nominal lifespan of the target load. , , , , , These are the weight coefficients for the first, second, third, fourth, fifth, and sixth health sub-items, respectively. It is an environmental correction factor.
[0012] In some embodiments of the first aspect of this application, after obtaining the health status of each target load, the method further includes: determining the health level of each target load based on each health status; wherein, the method of determining the health level of the target load includes: comparing the health status of the target load with a preset first health status threshold, a second health status threshold, and a third health status threshold to determine the health level of the target load; wherein, the first health status threshold is greater than the second health status threshold, and the second health status threshold is greater than the third health status threshold; if the health status of the target load is greater than or equal to the first health status threshold, then the health level of the target load is determined to be the first health status, indicating that the target load is in a good state; if the health status of the target load is less than the first health status threshold but greater than or equal to the second health status threshold, then the health level of the target load is determined to be the second health status, indicating that the target load is in a warning state; if the health status of the target load is less than the third health status threshold, then the health level of the target load is determined to be the fourth health status, indicating that the target load is in a critical state.
[0013] In some embodiments of the first aspect of this application, if the health level of the target load is the second health level, then based on a pre-built remaining lifetime prediction model of the target load, the remaining lifetime of the target load is predicted, and a first maintenance memorandum of the target load is generated and reported to the operation and maintenance platform, so that the operation and maintenance platform can generate a maintenance work order for the target load; if the health level of the target load is the third health level, then based on a pre-built remaining lifetime prediction model of the target load, the remaining lifetime of the target load is predicted, the fault type of the target load is identified, the spare parts required for the maintenance of the target load are determined, and a maintenance work order is generated. The second maintenance memorandum for the target load is reported to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load and the autonomous driving domain controller can adjust the vehicle task strategy of the autonomous logistics vehicle. If the health level of the target load is the fourth health level, the fault type of the target load is identified, the spare parts required for the maintenance of the target load are determined, and a third maintenance memorandum for the target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load and the autonomous driving domain controller can adjust the vehicle task strategy of the autonomous logistics vehicle.
[0014] In some embodiments of the first aspect of this application, the method for predicting the remaining lifetime of a target load includes: if the target load has a limited number of operations, constructing a corresponding remaining lifetime prediction model based on the cumulative number of switching operations of the target load to predict the current remaining available operations of the target load; if the health of the target load exhibits linear decay, obtaining multiple historical health values of the target load and the cumulative operating time corresponding to each historical health value, and performing linear fitting on the historical health values at each cumulative operating time to establish a trend model of the target load health value changing with operating time as its remaining lifetime prediction model to predict the current remaining operating time of the target load; if the health of the target load exhibits exponential decay, obtaining multiple historical health values of the target load and the cumulative operating time corresponding to each historical health value, and performing exponential fitting on the historical health values at each cumulative operating time to establish a trend model of the target load health value changing with operating time as its remaining lifetime prediction model to predict the remaining operating time of the target load.
[0015] In some embodiments of the first aspect of this application, the method for identifying the fault type of the target load includes: if the on-state voltage drop data at multiple consecutive sampling times in the original operating data of the target load is greater than 0, but the steady-state current value is 0, then the target load is determined to have an open-circuit fault; if the steady-state current value in the original operating data of the target load exceeds a preset short-circuit threshold of the nominal steady-state current, that is, the relative deviation of the steady-state current in the health feature data exceeds a preset short-circuit threshold, then the target load is determined to have a short-circuit fault; if the steady-state current value at multiple sampling times in the original operating data of the target load gradually decreases or fluctuates more, that is, the relative deviation of the steady-state current in the health feature data decreases or the steady-state current ripple rate increases, then the target load is determined to have an aging fault; if the steady-state current value at one or more mutually spaced sampling times in the original operating data of the target load is 0, then the target load is determined to have an intermittent fault; if the on-state voltage drop data at multiple consecutive sampling times in the original operating data of the target load is 0 and the steady-state current value is 0, then the load signal sampling circuit of the target load is determined to have a fault.
[0016] In some embodiments of the first aspect of this application, the method for determining the spare parts required for the maintenance of the target load includes: if the target load is replaceable, the spare parts required for maintenance are determined to be the standard spare parts of the target load; if the fault type of the target load is an intermittent fault, the spare parts required for maintenance are determined to be the connection structure in the target load; if the fault type of the target load is a load signal sampling circuit fault, the spare parts required for maintenance are determined to be the chassis domain controller maintenance spare parts.
[0017] To achieve the above and other related objectives, a second aspect of this application provides an unmanned logistics vehicle, comprising: a vehicle body and a chassis domain controller and an autonomous driving domain controller disposed on the vehicle body; wherein the chassis domain controller is connected to the autonomous driving domain controller and an operation and maintenance platform respectively, and is configured to execute the unmanned logistics vehicle load health management method as described in any of the above embodiments; the autonomous driving domain controller is connected to the operation and maintenance platform and is configured to adjust the vehicle task strategy of the unmanned logistics vehicle according to the maintenance memo of each target load.
[0018] To achieve the above and other related objectives, a third aspect of this application provides a cloud-edge collaborative platform for unmanned logistics vehicle load health management. The platform includes: multiple unmanned logistics vehicles, each including: a vehicle body and a chassis domain controller and an autonomous driving domain controller mounted on the vehicle body; wherein the chassis domain controller is connected to both the autonomous driving domain controller and an operation and maintenance platform, and is configured to execute the unmanned logistics vehicle load health management method as described in any of the above embodiments; the autonomous driving domain controller is connected to the operation and maintenance platform and is configured to adjust the vehicle task strategy of the unmanned logistics vehicle based on the maintenance memo for each target load; a cloud server is connected to each unmanned logistics vehicle and is configured to construct a health assessment model and a remaining life prediction model for each target load, and to acquire the original operating data, health characteristic data, and health status of each target load in real time, and periodically update each health assessment model and each remaining life prediction model.
[0019] As described above, this application has the following beneficial effects: It provides an unmanned logistics vehicle, a load health management method, and a cloud-edge collaborative platform. Through the chassis domain controller of the unmanned logistics vehicle, it collects raw operating data of multiple target loads in real time, thereby effectively utilizing the existing analog sampling and computing resources of the chassis domain controller to achieve non-intrusive load health monitoring. Based on multiple health assessment models, it differentiates the health of each target load, thereby effectively improving the adaptability and accuracy of health assessment. According to each health level, it executes predictive maintenance measures and hierarchical management for each target load, automatically identifies the fault type of the target load, accurately determines the spare parts required for its maintenance, generates a maintenance memo, and reports it to the operation and maintenance platform and the autonomous driving domain controller, thus forming a closed loop for load health management. This maximizes the operational efficiency of the unmanned logistics vehicle while ensuring its operational safety. Attached Figure Description
[0020] Figure 1 The diagram shown is a flowchart illustrating the load health management method for unmanned logistics vehicles in one embodiment of this application.
[0021] Figure 2 The diagram shown is a structural schematic of an unmanned logistics vehicle according to one embodiment of this application.
[0022] Figure 3 The diagram shown is a structural schematic of a cloud-edge collaborative platform for load health management of unmanned logistics vehicles, as described in one embodiment of this application. Detailed Implementation
[0023] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.
[0024] In the embodiments of this application, the terms "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect, without limiting their order. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" do not necessarily imply that they are different.
[0025] To address the problems mentioned above, this application provides an unmanned logistics vehicle, a load health management method, and a cloud-edge collaborative platform. The aim is to solve the technical problems of low fault location efficiency, inaccurate maintenance decisions, high maintenance costs, and high risk of unplanned downtime caused by the inability of existing technologies to timely and accurately identify early anomalies in the onboard load of unmanned logistics vehicles and assess their health status.
[0026] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application are further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining this application and are not intended to limit this application.
[0027] like Figure 1 The diagram shown illustrates a flowchart of a method for managing the load health of an unmanned logistics vehicle according to an embodiment of this application.
[0028] The unmanned logistics vehicle load health management method described in this embodiment is applied to the chassis domain controller of the unmanned logistics vehicle. The chassis domain controller is a centralized control unit located in the chassis system of the unmanned logistics vehicle. It is used to uniformly receive data from chassis-related sensors, actuators, and other domain controllers, and to coordinate and control the vehicle's driving, braking, steering, and suspension functions. The chassis domain controller integrates multiple analog sampling channels and connects to various sensors. It can utilize the built-in current sampling, voltage sampling, and temperature sampling functions of the HSD (High Side Driver) chip in each analog sampling channel to periodically collect actual operating data of various loads of the unmanned logistics vehicle monitored by each sensor. Simultaneously, the chassis domain controller also includes a processor, memory, and communication interface, possessing certain computing capabilities. Therefore, this application utilizes the existing analog sampling and computing resources of the chassis domain controller, without the need for a new independent controller, to achieve a non-intrusive, low-cost, and highly integrated unmanned logistics vehicle load health management solution.
[0029] The chassis domain controller of the unmanned logistics vehicle is connected to the autonomous driving domain controller, which can send the unmanned logistics vehicle load health management solution to the autonomous driving domain controller. Based on this, the autonomous driving domain controller can make vehicle task decisions, vehicle path planning and vehicle driving decisions, forming a closed loop of load health management application.
[0030] The chassis domain controller of the unmanned logistics vehicle is also connected to an external operation and maintenance platform. It can send the unmanned logistics vehicle load health management solution to the operation and maintenance platform, so that operation and maintenance personnel can manually control the unmanned logistics vehicle to drive and operate safely through the operation center, further enriching the application closed loop of load health management.
[0031] The unmanned logistics vehicle load health management method in this embodiment is used to assess the health status of multiple target loads of the unmanned logistics vehicle and identify possible early anomalies. These target loads include, but are not limited to, loads installed on the unmanned logistics vehicle such as lights, horns, fans, water pumps, solenoid valves, electronic locks, chillers, motors, and heaters. Specifically, the unmanned logistics vehicle load health management method mainly includes the following steps, such as... Figure 1 As shown.
[0032] Step S1: Collect raw operating data of multiple target loads in the unmanned logistics vehicle in real time, and extract features from each raw operating data to generate health feature data for each target load.
[0033] Specifically, a time window is defined, such as a 1-minute time window, and this time window is slid based on a preset step size. Raw operating data generated by the target load within each time window (e.g., within a consecutive 1-minute period) is collected in real time. Feature extraction is performed on each raw operating data point to generate corresponding health feature data for each time window (e.g., within each 1-minute period), thereby achieving real-time monitoring and updating of the target load's health status at a high frequency. In a preferred embodiment, the sampling frequency of the raw operating data can be set to no less than 100Hz (i.e., one data point is sampled every 10 milliseconds), ensuring that the surge current peak and back EMF drop waveform at the moment of target load startup can be captured.
[0034] In one embodiment, the chassis domain controller of the unmanned logistics vehicle is equipped with an independent load signal sampling path for each target load. The original operating data of the corresponding target load can be continuously collected by each load signal sampling path, thereby reducing sampling interference between different target loads and sampling delay caused by channel multiplexing. This establishes an accurate correspondence between the target load and the original operating data, improves the ability to capture transient operating characteristics and the accuracy and real-time performance of the original operating data, and provides a reliable data foundation for the extraction of health features, health assessment and fault location of each target load.
[0035] The original operating data includes at least one or more combinations of the following: on / off event monitoring data of the corresponding target load within the current time window, continuous operation monitoring data, and environmental compensation data.
[0036] The switch event monitoring data includes at least one or more combinations of the following: the number of switch-on events corresponding to the target load within the current time window, the peak value and duration (in milliseconds) of the surge current generated by each switch-on event, the number of switch-off events, and the back EMF drop waveform for each switch-off event. The back EMF drop waveform reflects the back EMF waveform characteristics generated after the corresponding target load is de-energized, especially the back EMF waveform characteristics of inductive loads (such as motors and solenoid valves), including the back EMF amplitude at multiple sampling times after power failure.
[0037] The continuous operation monitoring data includes at least: the power-on time of the corresponding target load within the current time window, and the steady-state current value and on-state voltage drop data at multiple sampling times. In one embodiment, the on-state voltage drop data can be represented by the voltage difference across the corresponding target load when it is in the on-state, and can be specifically estimated based on the product of the on-resistance of the HSD chip in the chassis domain controller and the steady-state current value.
[0038] The environmental compensation data includes at least: ambient temperature and battery voltage. The ambient temperature can be acquired by the onboard NTC (Negative Temperature Coefficient Thermistor) of the chassis domain controller. The battery voltage is the voltage provided by the battery of the autonomous logistics vehicle and can be used as a voltage reference to normalize the acquired current values. In one embodiment, the sampling frequency of the environmental compensation data can be appropriately reduced to 1Hz.
[0039] Furthermore, after obtaining the original operating data of each target load, feature data used to characterize the health status of the target load are extracted from each original operating data to obtain the health feature data of each target load within the current time window.
[0040] In one embodiment, the health characteristic data includes at least one or more combinations of the following: surge current health characteristic data of the corresponding target load within the current time window, steady-state current health characteristic data, on-state voltage drop health characteristic data, cumulative switching count, and cumulative running time.
[0041] The cumulative number of switches is the sum of the number of switch-on events and switch-off events accumulated since the corresponding target load was put into use. The cumulative operating time is the sum of the cumulative power-on time of the corresponding target load accumulated since it was put into use, in hours (h).
[0042] In one embodiment, the surge current health characteristic data includes at least: the surge current impact coefficient and the maximum surge current impact coefficient for each switch-on event of the target load within the current time window.
[0043] Specifically, the formula for calculating the surge current impact factor is as follows:
[0044] ;Formula (1)
[0045] in, For the corresponding target load within the current time window The surge current impact factor of the secondary switch opening event; For the corresponding target load within the current time window The peak value of the surge current generated by the next switch opening event, in amperes (A). For the corresponding target load within the current time window The average steady-state current during the energization period after the switch is turned on, in amperes (A). , For the corresponding target load in the first During the power-on period after the secondary switch is turned on The steady-state current value at the sampling time is expressed in amperes (A). In a specific embodiment, the steady-state current value can be determined when the rate of change of current at multiple consecutive sampling times (such as 10 sampling times) is less than ±5%.
[0046] In this embodiment, the maximum surge current impact coefficient is the maximum value of the surge current impact coefficients of the corresponding target load within the current time window. This is used to more sensitively detect early signs of target load performance degradation (such as an abnormally large surge during startup) when the target load may experience multiple on / off events (such as multiple flashing of the turn signal), avoiding the masking of individual abnormal events by taking an average value. The formula for calculating the maximum surge current impact coefficient is:
[0047] ;Formula (2)
[0048] in, This is the maximum surge current impact factor for the target load within the current time window.
[0049] It should be noted that the surge current impact factor of resistive loads (such as lights, heaters, etc.) is generally less than 1.5 when operating normally. When the surge current impact factor increases to 2 or higher, it indicates that the resistive load may have aging components or poor contact, posing a health risk. The surge current impact factor of inductive loads (such as fans, chillers, motors, etc.) is generally 5-8 when operating normally. When the surge current impact factor increases to 10 or higher, it indicates that the inductive load may have health risks such as bearing jamming or short circuits between winding turns.
[0050] In one embodiment, the steady-state current health characteristic data includes at least: the relative deviation of steady-state current at multiple sampling times corresponding to the target load within the current time window, and one or more combinations of the steady-state current average value, steady-state current peak-to-peak value, steady-state current ripple rate, and steady-state current average relative deviation within the current time window.
[0051] Specifically, the average steady-state current is equal to the average of the steady-state current values of the corresponding target load within the current time window, that is:
[0052] ;Formula (3)
[0053] in, The average steady-state current of the target load within the current time window is expressed in amperes (A). For the corresponding target load within the current time window The steady-state current value at the sampling time, in amperes (A).
[0054] The peak-to-peak value of the steady-state current is used to reflect the fluctuation range of each steady-state current value within the current time window. It can be represented by the difference between the maximum and minimum values of the steady-state current of the corresponding target load within the current time window, i.e.:
[0055] ;Formula (4)
[0056] in, The peak-to-peak steady-state current of the target load within the current time window is expressed in amperes (A).
[0057] The steady-state current ripple rate is used to reflect the wear condition of key components in the corresponding target load, such as whether the motor commutator, brushes, and bearings are worn or whether the load is uneven. The specific calculation formula is as follows:
[0058] ;Formula (5)
[0059] in, The steady-state current ripple rate of the target load within the current time window is expressed as a percentage (%).
[0060] The formula for calculating the relative deviation of the steady-state current is:
[0061] ;Formula (6)
[0062] in, For the corresponding target load within the current time window The relative deviation of steady-state current at the sampling time, expressed as a percentage (%). For the corresponding target load within the current time window The steady-state current value at the sampling time, in amperes (A); The nominal steady-state current corresponding to the target load is expressed in amperes (A).
[0063] The formula for calculating the average relative deviation of the steady-state current is:
[0064] ;Formula (7)
[0065] in, This represents the average relative deviation of the steady-state current of the target load within the current time window, expressed as a percentage (%). It should be noted that this average relative deviation of the steady-state current better reflects the true current characteristics of the target load within the current time window and has strong noise immunity. When the average relative deviation of the steady-state current is positive and excessively large, it may indicate that the target load is overloaded or is a potential precursor to a short-circuit fault. When the average relative deviation of the steady-state current is negative and excessively large, it may indicate that the target load may have an open-circuit fault or abnormal power supply voltage, among other health risks.
[0066] In one embodiment, the conduction voltage drop health characteristic data includes at least the conduction voltage drop trend of the corresponding target load within the current time window. Specifically, the conduction voltage drop trend can be represented by the slope of the conduction voltage drop changing over time. Specifically, the conduction voltage drop data at each sampling time can be linearly fitted, and the slope of the fitted line can be determined as the conduction voltage drop trend. If the conduction voltage drop trend is consistently greater than 0, i.e., the conduction voltage drop data gradually increases over time, it may indicate that the target load has health risks such as oxidation of the circuit or connector, or aging of the HSD chip. In one embodiment, the unit of the conduction voltage drop trend can be... sky.
[0067] If the target load is an inductive load, the health characteristic data further includes: back EMF decay health characteristic data. This back EMF decay health characteristic data includes at least: the back EMF decay time constant and the maximum back EMF decay time constant for each switch-off event of the target load within the current time window.
[0068] Specifically, the back EMF decay time constant is used to reflect the change trend of the back EMF over time after the target load of inductive type is de-energized. For inductive loads such as motors and solenoid valves, the back EMF after de-energization can be approximately expressed as an exponential decay, and the formula for calculating the back EMF decay time constant is as follows:
[0069] ;Formula (8)
[0070] in, To correspond to the target load within the current time window After the secondary switch is closed and power is cut off The time constant of the decay of the back electromotive force at the sampling time; To correspond to the target load within the current time window After the secondary switch is closed and power is cut off The amplitude of the back electromotive force at the sampling time, in volts (V). To correspond to the target load within the current time window After the secondary switch is closed and power is cut off The amplitude of the back electromotive force at the sampling time, in volts (V). for Sampling time and The duration between sampling moments, in seconds (s).
[0071] In this embodiment, the maximum back EMF decay time constant is the maximum value of the back EMF decay time constants for the corresponding target load within the current time window, that is:
[0072] ;Formula (9)
[0073] in, This is the time constant for the decay of the maximum back EMF of the target load within the current time window.
[0074] It should be noted that if the back electromotive force (EMF) decays more slowly after the power is cut off to an inductive target load, i.e., the back EMF decay time constant decreases, it may indicate a health risk to the target load. For example, if the back EMF decays more slowly after the power is cut off to a motor, it indicates an increase in bearing resistance.
[0075] Step S2: Based on the pre-built health assessment model for each target load, assess the health of each target load according to the health characteristic data.
[0076] In one embodiment, a corresponding health assessment model is pre-constructed for each target load of the autonomous logistics vehicle. Each health assessment model can employ the same mathematical model framework, and its parameters are configured according to the load type and characteristics of the corresponding target load to adapt to different target loads. For example, different weighting coefficients can be configured for the mathematical model framework for different target loads to characterize the different degrees of influence of various health characteristic data on the health status of the corresponding target load. This achieves differentiated health assessments for target loads of different types and characteristics, improving the adaptability and accuracy of the health assessment results with the actual operating status of each target load.
[0077] In this embodiment, the same mathematical model framework used by each health assessment model can be expressed as:
[0078] ;Formula (10)
[0079] in, The health score of the target load is used to characterize the health status of the target load up to the current time window. It is expressed on a 100-point scale. When the health score is 100, it indicates that the target load is currently in a fully healthy state; when the health score is 0, it indicates that the target load is currently completely unresponsive.
[0080] This is the first health sub-item for the target load, used to characterize the health level of the target load, indicating the deviation of its steady-state current value from its nominal steady-state current value within the current time window. In one embodiment, this can be based on the average relative deviation of the target load's steady-state current within the current time window. The ratio between the allowable current offset of the target load and the target load determines the first health sub-item. When the average relative deviation of the steady-state current When the deviation from the allowable current offset is within ±10%, the first health sub-item... It is 100 points; when the average relative deviation of the steady-state current is... When the deviation from the allowable current offset is between 10% and 30%, the first health sub-item... The linearity drops to sixty points; when the average relative deviation of the steady-state current... When the deviation from the allowable current offset exceeds 30%, the first health sub-item... The score dropped linearly to zero.
[0081] This is the second health sub-item for the target load, used to characterize the health level of the target load as indicated by the surge current impact coefficient within the current time window. In one embodiment, the surge current impact coefficient of the target load within the current time window can be used as the indicator. The ratio between the surge current impact factor and the allowable surge current impact factor of the target load determines the second health sub-item. When there are multiple switch-on events within the current time window, the maximum surge current impact coefficient of the target load within the current time window can be used as the criterion. The second health sub-item is determined by the ratio between the allowable surge current impact factor of the target load and the target load. When the maximum surge current impact coefficient When the surge current is less than 1.2 times the allowable surge current impact factor, the second health sub-item... It is 100 points; when the maximum surge current impact coefficient is... When the allowable surge current impact factor is between 1.2 and 2.0 times, the second health sub-item The linear reduction to fifty points; when the maximum surge current impact factor When the surge current exceeds 2.0 times the allowable surge current impact factor, the second health sub-item... The score dropped linearly to zero.
[0082] This is the third health sub-item for the target load, used to characterize the health level of the target load as indicated by the steady-state current fluctuation amplitude within the current time window. In one embodiment, it can be based on the peak-to-peak value of the steady-state current of the target load within the current time window. The third health sub-item is determined by the ratio between the allowable current fluctuation range of the target load and the target load. When the peak-to-peak value of the steady-state current... When the deviation from the allowable current fluctuation range is less than 5%, the third health sub-item It is 100 points; when the peak-to-peak value of the steady-state current is... When the deviation from the allowable current fluctuation range is between 5% and 15%, the third health sub-item The linearity drops to fifty points; when the peak-to-peak value of the steady-state current... When the deviation from the allowable current fluctuation range exceeds 15%, the third health sub-item... The score dropped linearly to zero.
[0083] This is the fourth health sub-item for the target load, used to characterize the health level of the target load as indicated by the on-state voltage drop trend within the current time window. In one embodiment, the fourth health sub-item can be determined based on the on-state voltage drop trend of the target load within the current time window. When the on-state voltage drop trend is less than 0, the fourth health sub-item... It is 100 points; when the on-state voltage drop trend is between 0 and 50... Between days, the fourth health sub-item A linear drop to 60 indicates a slight increase in the on-state voltage drop trend, requiring close observation; when the on-state voltage drop trend is between 50 and 200... Between days, the fourth health sub-item A linear drop to 30 indicates a significant increase in the on-state voltage drop, requiring a warning; when the on-state voltage drop exceeds 200... Timing, the fourth health sub-item A linear drop to zero indicates a severely increased on-state voltage drop, placing the target load in a critical condition. Increased on-state voltage drop is typically caused by oxidation of line connectors, aging of relay contacts, or increased internal resistance of the HSD chip. The above is used to determine the fourth health sub-item. The threshold values are preferred values set based on the experience values of typical vehicle electrical systems. This application does not specifically limit them. Users can calibrate and adjust them according to the specific unmanned logistics vehicle model and the corresponding target load characteristics in actual applications.
[0084] This is the fifth health sub-item for the target load, used to characterize the health level of the target load as indicated by the back electromotive force decay characteristics within the current time window for inductive type target loads. In one embodiment, the fifth health sub-item can be determined based on the changing trend of the back electromotive force decay time constant of the target load within the current time window. When there are multiple switch-off events within the current time window, the fifth health sub-item can be determined based on the changing trend of the maximum back EMF decay time constant of the target load within the current time window. The trend of the back EMF decay time constant can be represented by the slope of its change over time, and the trend of the maximum back EMF decay time constant can also be represented by the slope of its change over time. When the bearing of the inductive target load is worn or poorly lubricated, the mechanical resistance increases, the speed drops faster after power failure, the back EMF decays more rapidly, and the maximum back EMF decay time constant decreases. When the trend of the maximum back EMF decay time constant is less than 0.1 ms / day, the fifth health sub-item... The score is 100; when the variation trend of the maximum back electromotive force decay time constant is between 0.1 and 0.5 ms / day, the fifth health sub-item... The score linearly decreases to fifty points; when the variation trend of the maximum back electromotive force decay time constant is between 0.5 and 2.0 ms / day, the fifth health sub-item... The linear rate decreases to 10; when the change trend of the maximum back electromotive force decay time constant exceeds 2.0 ms / day, the fifth health sub-item... The linearity drops to zero. The back EMF decay time constant reflects the rotational inertia and mechanical resistance of inductive loads such as motor rotors. In the early stages of bearing wear, the resistance increases slowly, and the back EMF decay time constant changes by 0.1-0.5 ms per day; when wear is severe, the resistance increases sharply, and the back EMF decay time constant changes by more than 0.5 ms per day; when the bearing is about to seize, the back EMF decay time constant changes by more than 2 ms per day. The above is used to determine the fifth health sub-item. The threshold values are preferred values based on experience with typical vehicle motors. This application does not specifically limit them, and users can calibrate and adjust them according to the specific target load type in actual applications.
[0085] This is the sixth health sub-item for the target load, used to characterize the degree to which the cumulative number of switching operations or cumulative on-time of the target load consumes its service life. In one embodiment, the sixth health sub-item can be determined based on the percentage between the cumulative number of switching operations or cumulative on-time and the rated operating life of the target load. .
[0086] , , , , , These are the weighting coefficients for the first, second, third, fourth, fifth, and sixth health sub-items, respectively. Different weighting coefficients are used when calculating the health of different target loads. For example, the weighting coefficient for the second health sub-item of the lighting load... The weighting coefficient of the third health sub-item of motor load is high. High. The weighting coefficients for each target load can be obtained and updated in the cloud server using a pre-built health assessment parameter prediction model.
[0087] This refers to the environmental correction factor. The environmental correction factor varies under different ambient temperatures and battery voltages. For example, the environmental correction factor will decrease when the ambient temperature is too high or the battery voltage is too low. This will decrease, thus reducing the health of the target load. It will also decrease. In one embodiment, ;in, For ambient temperature correction factor, This is the voltage correction factor. The ambient temperature correction factor... and the voltage correction factor It can be obtained by using a lookup table and interpolation method based on a pre-built environmental compensation parameter mapping table.
[0088] In one embodiment, the nominal values of each target load, such as nominal steady-state current, allowable surge current impact factor, allowable current fluctuation amplitude, allowable current offset, and rated operating life (e.g., rated number of operations or rated usage time), can be pre-calibrated and a nominal parameter table can be constructed and stored in the chassis domain controller of the autonomous vehicle. Maintenance personnel can calibrate and update each nominal value through a cloud server or diagnostic tools. When calibrating each nominal value, the influence of ambient temperature and battery voltage must be considered. For example, the allowable current offset at high temperatures can be appropriately relaxed. The nominal parameter table includes at least one nominal value for each target load under different ambient temperatures and battery voltages.
[0089] The weight coefficients for each target load can be obtained and updated in the cloud server using a pre-built health assessment parameter prediction model.
[0090] Step S3: Based on each health status, predict the remaining lifespan of each target load according to the pre-built remaining lifespan prediction model of each target load, identify the fault type of each target load, determine the spare parts required for maintenance of each target load, generate a maintenance memo for each target load and report it to the operation and maintenance platform and the autonomous driving domain controller.
[0091] In one embodiment, after obtaining the health status of each target load, the health level of each target load can be further determined.
[0092] Specifically, the health score of the target load is compared with preset first health score thresholds, second health score thresholds, and third health score thresholds to determine the health level of the target load. The first health score threshold is greater than the second health score threshold, and the second health score threshold is greater than the third health score threshold. In a preferred embodiment, the first health score threshold can be preset to 80 points, the second health score threshold can be preset to 60 points, and the third health score threshold can be preset to 40 points.
[0093] If the health of the target load is greater than or equal to the first health threshold (e.g.) If the score is 80, the target load is determined to be in the first health level, indicating that the target load is in good condition. At this time, normal monitoring of the target load is sufficient, and no maintenance measures are required.
[0094] If the health score of the target load is less than the first health score threshold and greater than or equal to the second health score threshold (e.g., 60 points), If the score is 80, the health level of the target load is determined to be the second health level, indicating that the target load is in a state of concern. At this time, the health characteristic data of the target load show slight deviations, and the target load shows an aging trend. It can be inspected and maintained during the next regular maintenance.
[0095] If the health score of the target load is less than the second health score threshold and greater than or equal to the third health score threshold (e.g., 40 points), If the score is 60, the health level of the target load is determined to be the third health level, indicating that the target load is in a warning state. At this time, the health characteristic data of the target load deviates significantly, and failure may occur in the near future (such as within a week). It is recommended to proactively warn the operation and maintenance personnel and arrange planned inspections and maintenance in a timely manner.
[0096] If the health of the target load is less than the third health threshold (e.g.) If the score is 40 points, the health level of the target load is determined to be the fourth health level, indicating that the target load is in a critical state. At this time, the health characteristic data of the target load have deviated significantly, and the target load may fail at any time. It is recommended to stop using the target load immediately, and to alert the operation and maintenance personnel to arrange for emergency inspection and maintenance.
[0097] In one embodiment, the health level of the target load can also be determined based on its health degradation rate. Specifically, if the health degradation rate of the target load exceeds a preset health degradation threshold multiple times consecutively, the health level of the target load is determined to be level four, indicating that the target load is deteriorating rapidly and its use should be stopped immediately. An alarm should be triggered to the maintenance personnel, and emergency repairs and maintenance should be arranged. The health degradation rate can be determined based on the daily health degradation value. In this case, the health degradation threshold can be set to 2 points / day. If the health degradation value of the target load exceeds 2 points per day for several consecutive days (e.g., 3 consecutive days), the health level of the target load is determined to be level four.
[0098] Once the health level of each target load is determined, predictive maintenance measures can be performed on each target load according to its health level. This includes predicting the remaining service life of the target load based on its health degradation trend and distinguishing between health levels. When the health status of a target load is "attention, warning, or critical", it supports hierarchical management of planned maintenance and emergency handling for the target load.
[0099] In one embodiment, if the health level of the target load is the second health level, the remaining lifespan of the target load is predicted based on the pre-built remaining lifespan prediction model of the target load, and a first maintenance memorandum of the target load is generated and reported to the operation and maintenance platform so that the operation and maintenance platform can generate a maintenance work order for the target load.
[0100] Specifically, different types of target loads can use different remaining lifetime prediction models. A corresponding remaining lifetime prediction model can be pre-built for each target load of the autonomous logistics vehicle to predict the current remaining lifespan of each target load.
[0101] For target loads with a clearly defined limit on the number of operations (such as relay contacts, solenoid valves, electronic locks, etc.), a corresponding remaining lifetime prediction model can be constructed based on the cumulative number of switching operations of the target load to predict the current remaining usable operations of the target load. Specifically, the remaining lifetime prediction model for the target load can be expressed as:
[0102] ;Formula (11)
[0103] in, The number of times the target load is currently available. The rated number of actions for the target load; This represents the cumulative number of times the device has been switched on and off up to the current time. The current health status of the target load is expressed as a percentage. As the rated number of actions A health-based correction factor is used to adjust the rated number of actions for the target load. .
[0104] For target loads with obvious aging trends and linear decline in health (such as lights, horns, heaters, etc.), a corresponding remaining life prediction model can be constructed based on the linear health decline characteristics of the target load to predict the current remaining operating time of the target load.
[0105] Specifically, depending on the load type of the target load, if the target load's health status declines linearly, multiple historical health statuses and the corresponding cumulative running time for each historical health status are obtained. A linear fit is then performed on the historical health status at each cumulative running time to establish a trend model of the target load's health status changing with running time, which serves as its remaining lifetime prediction model. For example, health status change data for the target load over the past thirty days is obtained, and a linear fit is performed to obtain the remaining lifetime prediction model for the target load.
[0106] The remaining lifetime prediction model can be expressed as:
[0107] ;Formula (12)
[0108] in, Accumulated operation for target load The estimated health status based on duration is expressed on a 100-point scale. To characterize the slope of health status as a function of operating time, This indicates that the health of the target load decreases over time. This is the intercept.
[0109] Furthermore, based on this remaining lifetime prediction model, the estimated runtime of the target load at the third health threshold (e.g., health score of 40) is obtained, and the difference between this estimated runtime and the cumulative runtime of the target load up to the present is taken as the current remaining runtime of the target load. The formula for calculating the remaining runtime is as follows:
[0110] ;Formula (13)
[0111] in, The remaining runtime of the target load, in hours (h). The current cumulative running time of the target load, in hours (h). In this embodiment, the estimated runtime of the target load at the third health threshold (e.g., health score of 40) can be calculated using formula (12) to serve as the runtime, i.e. , The third health threshold is preferably 40 points.
[0112] For target loads whose health declines exponentially (such as motors, fans, water pumps, chillers, etc.), a corresponding remaining life prediction model can be constructed based on the exponential health decline characteristics of the target load to predict the current remaining operating time of the target load.
[0113] Specifically, based on the load type of the target load, for target loads with exponentially declining health, multiple historical health values and the corresponding cumulative runtimes for each historical health value are obtained. An exponential fit is then performed on the historical health values at each cumulative runtime to establish a trend model of the target load's health value changing with runtime, which serves as its remaining lifetime prediction model. This remaining lifetime prediction model can be expressed as:
[0114] ;Formula (14)
[0115] in, and These are the exponential parameters obtained from exponential fitting. This indicates that the health of the target load decreases over time.
[0116] Furthermore, based on the remaining lifetime prediction model, the estimated running time corresponding to the third health threshold (e.g., health score of 40) for the target load is obtained, and the difference between the estimated running time and the cumulative running time of the target load up to the present is taken as the current remaining running time of the target load. The formula for calculating the remaining running time can be found in formula (13). In this embodiment, .
[0117] In one embodiment, the remaining lifetime prediction model for each target load can be trained and updated in a cloud server.
[0118] After predicting the remaining useful life of one or more target loads at the second health level, a first maintenance memorandum can be generated for each target load. The first maintenance memorandum includes at least: the health status, health level, remaining useful life, and recommended maintenance time for the corresponding target load. The remaining useful life includes: the remaining number of available uses or the remaining operating time of the target load. The recommended maintenance time is the scheduled maintenance time for the target load.
[0119] In one embodiment, the first maintenance memo for each target load at the second health level can be reported to the operation and maintenance platform, allowing the platform to generate maintenance work orders for each target load. This enables operation and maintenance personnel to schedule routine inspections and maintenance of each target load, avoiding unplanned downtime of the autonomous logistics vehicles and improving their operational efficiency. The maintenance work order includes at least: the health status, health level, remaining service life, planned maintenance time, and maintenance priority of the corresponding target load. The planned maintenance time may be equal to the suggested maintenance time. The maintenance priority can be determined based on the health level of the target load and includes: a first maintenance priority, a second maintenance priority, and a third maintenance priority. Different levels of maintenance priority characterize the urgency of the target load's maintenance, with the first maintenance priority being the most urgent, the second priority allowing for some de-escalation, and the third priority being more lenient. Specifically, if the health level of the target load is the second health level, then the maintenance priority of the target load is the third maintenance priority; if the health level of the target load is the third health level, then the maintenance priority of the target load is the second maintenance priority; if the health level of the target load is the fourth health level, then the maintenance priority of the target load is the first maintenance priority.
[0120] When the health level of the target load is the second health level, the target load will not fail immediately and can be inspected and maintained during regular maintenance. Therefore, in a preferred embodiment, the first maintenance memorandum of the target load can be reported to the operation and maintenance platform on a regular basis, such as at a specified sampling time on the same day or after the target load is powered off on the same day, without having to report in real time for every time window, thereby reducing unnecessary data transmission and processing.
[0121] In one embodiment, if the health level of the target load is the third health level, the remaining lifespan of the target load is predicted based on the pre-built remaining lifespan prediction model of the target load, the fault type of the target load is identified, the spare parts required for the maintenance of the target load are determined, and a second maintenance memorandum of the target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load and the autonomous driving domain controller can adjust the vehicle task strategy of the unmanned logistics vehicle.
[0122] Specifically, if the health level of the target load is the third health level, the remaining lifetime prediction model and remaining lifetime calculation method shown in formulas (11)-(14) can be used to calculate the remaining available times or remaining running time for different types of target loads. For the sake of simplicity, it will not be elaborated here.
[0123] After obtaining the remaining service life of one or more target loads at the third health level, the method further includes: automatically identifying the fault type of each target load based on the original operating data and health characteristic data of each target load, and determining whether each target load is an open circuit fault, short circuit fault, aging fault, intermittent fault, or load signal sampling circuit fault.
[0124] Specifically, if the on-state voltage drop data of the target load is greater than 0 at multiple consecutive sampling times in the original operating data, but the steady-state current value is 0, then the target load is determined to have an open-circuit fault. Preferably, if the on-state voltage drop data of the target load is greater than 0 for 0.5s or more, but the steady-state current value is 0, then the target load is determined to have an open-circuit fault.
[0125] If any steady-state current value in the original operating data of the target load exceeds a preset short-circuit threshold of the nominal steady-state current, that is, if any steady-state current relative deviation in the health characteristic data exceeds the preset short-circuit threshold, then the target load is determined to have a short-circuit fault. Preferably, if the steady-state current value of the target load instantaneously exceeds 5 times or more of the nominal steady-state current, then the target load is determined to have a short-circuit fault.
[0126] If the steady-state current value at multiple sampling times in the original operating data of the target load gradually decreases or fluctuates more, that is, the relative deviation of the steady-state current in the health characteristic data decreases or the steady-state current ripple rate increases, then the target load is determined to have an aging fault.
[0127] If the steady-state current value is 0 at one or more mutually spaced sampling times in the original operating data of the target load, that is, the steady-state current value is briefly interrupted or fluctuates, but can recover within a specified time, then the target load is determined to have an intermittent fault, such as loose connectors or relay contact jitter.
[0128] If the on-state voltage drop data and steady-state current value are 0 at multiple consecutive sampling times in the original operating data of the target load, it is determined that the load signal sampling circuit of the target load has failed, such as the corresponding HSD chip in the chassis domain controller being damaged.
[0129] Furthermore, after completing the fault identification of one or more target loads at the third health level, the process also includes: determining the spare parts required for the maintenance of each target load based on the fault type of each target load, thereby realizing automatic fault diagnosis and precise spare parts delivery for the target loads.
[0130] In one embodiment, for replaceable target loads (such as light bulbs, motors, and solenoid valves), the spare parts required for maintenance can be directly identified as standard spare parts for that target load. For target loads with intermittent faults, the spare parts required for maintenance can be identified as connection structures within that target load, such as wiring harnesses or connectors. For target loads with load signal sampling circuit faults, the spare parts required for maintenance can be identified as spare parts for the chassis domain controller.
[0131] Based on the remaining lifespan, fault type, and required spare parts for each target load, a second maintenance memorandum is generated for each target load. This second maintenance memorandum includes at least: the health status, health level, remaining lifespan, recommended maintenance time, fault type, and required spare parts for maintenance of the corresponding target load. The remaining lifespan includes the remaining number of available cycles or remaining operating time for the target load. The recommended maintenance time is within one week and can be adaptively adjusted based on the specific health status value. The lower the health status, the earlier the recommended maintenance time.
[0132] In one embodiment, the second maintenance memo for each target load at the third health level can be reported to the operation and maintenance platform. This allows the platform to generate maintenance work orders for each target load, enabling maintenance personnel to schedule planned maintenance and repairs for each target load as early as possible. This avoids unplanned downtime of autonomous logistics vehicles, improves their operational efficiency, and forms a closed-loop application for load health management. In this case, the maintenance work order has the second maintenance priority, and predictive maintenance measures for the target load can be implemented. Furthermore, the maintenance work order can also include the corresponding fault type and required spare parts for the target load. This allows the operation and maintenance platform to match the nearest maintenance service point to the autonomous logistics vehicle based on its delivery task range and reminds the service point to pick and prepare the necessary parts in advance, achieving precise work assignment and improving the maintenance efficiency of the target load.
[0133] When the health level of the target load is the third health level, the target load will not fail immediately and can be scheduled for inspection and maintenance. Therefore, in a preferred embodiment, the second inspection memo of the target load can be reported to the operation and maintenance platform regularly based on a preset warning frequency, such as every 2 hours, so as to intensively monitor the health changes of the target load and be alert to the accelerated deterioration of the target load.
[0134] In one embodiment, the second maintenance memorandum for each target load of the third health level can also be reported to the autonomous driving domain controller of the autonomous logistics vehicle, so that the autonomous driving domain controller can adjust the vehicle mission strategy of the autonomous logistics vehicle.
[0135] Specifically, if the target load of the third health level is a non-safety-critical load, such as an entertainment screen or auxiliary lighting, the autonomous driving domain controller can ignore the health risk of the target load, keep the vehicle mission strategy unchanged, and control the unmanned logistics vehicle to continue to perform delivery tasks normally.
[0136] If the target load for the third health level is a safety-critical load, such as brake lights or turn signals, the autonomous driving domain controller adjusts the vehicle task strategy, reducing the vehicle speed of the autonomous logistics vehicle (e.g., from 40 km / h to 25 km / h), and plans for the autonomous logistics vehicle to travel to the maintenance service point recommended by the operation and maintenance platform according to the suggested maintenance time, so that the autonomous logistics vehicle can be inspected and maintained at the maintenance service point. At the same time, the autonomous driving domain controller can control the autonomous logistics vehicle to reduce the use of this target load, especially avoiding the use of this target load on highways or complex road sections.
[0137] In this embodiment, the autonomous driving domain controller can adaptively adjust the vehicle task strategy of the unmanned logistics vehicle according to the health level of each target load, such as performing delivery tasks normally, driving at reduced speed to a maintenance service point, or making an emergency safe stop, thereby realizing the collaborative scheduling of the chassis domain controller and the autonomous driving domain controller, forming an application closed loop of load health management, and maximizing operational efficiency while ensuring safety.
[0138] In one embodiment, if the health level of the target load is the fourth health level, the fault type of the target load is identified, the spare parts required for the maintenance of the target load are determined, a third maintenance memorandum for the target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load and the autonomous driving domain controller can adjust the vehicle task strategy of the unmanned logistics vehicle.
[0139] It should be noted that if the health level of the target load is level four, it indicates that the target load is in a critical state and may fail at any time. In this case, the remaining service life of the target load is determined to be 0, and the recommended maintenance time is the current moment. At the same time, the fault type of the target load must be identified immediately, the spare parts required for maintenance must be determined, a third maintenance memo for the target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller of the unmanned logistics vehicle, thereby alerting the operation and maintenance personnel, arranging emergency maintenance, and immediately stopping the use of the target load.
[0140] It should also be noted that the fault type identification method and the spare parts required for maintenance of the fourth health level target load can be adopted using the fault type identification method and spare parts determination method for the third health level target load provided in the above embodiments. For the sake of simplicity, they will not be described in detail here.
[0141] Furthermore, after obtaining the remaining service life, fault type, and required spare parts for maintenance of one or more target loads at the fourth health level, the process further includes: generating a third maintenance memorandum for each target load. The third maintenance memorandum includes at least: the health status, health level, remaining service life, recommended maintenance time, fault type, and required spare parts for maintenance of the corresponding target load. The remaining service life is 0; the recommended maintenance time is the current time.
[0142] In one embodiment, the third maintenance memorandum for each target load at the fourth health level should be immediately reported to the operation and maintenance platform. This allows the platform to generate maintenance work orders for each target load, enabling maintenance personnel to promptly inspect and maintain them, preventing unplanned downtime of the autonomous logistics vehicle and improving its operational efficiency. In this case, the maintenance work order has the highest priority, requiring immediate emergency maintenance measures for the target load. For example, the platform can notify maintenance personnel to go to the autonomous logistics vehicle site to replace or repair the target load. Furthermore, the maintenance work order can include the corresponding fault type and required spare parts for the target load. This allows the operation and maintenance platform to match the nearest safe parking point for the autonomous logistics vehicle based on its current location and remind maintenance personnel to bring the required spare parts when going to the safe parking point, achieving "accurate diagnosis, precise spare parts, and one-time repair," thus improving the maintenance efficiency of the target load.
[0143] In one embodiment, the third maintenance memorandum of each target load at the fourth health level can also be reported to the autonomous driving domain controller of the autonomous logistics vehicle in a timely manner, so that the autonomous driving domain controller can adjust the vehicle task strategy of the autonomous logistics vehicle to further maximize operational efficiency and ensure the safety of the autonomous logistics vehicle.
[0144] Specifically, if the target load of the fourth health level is a non-safety-critical load, such as an entertainment screen or auxiliary lighting, the autonomous driving domain controller can ignore the health risk of the target load, keep the vehicle mission strategy unchanged, and control the unmanned logistics vehicle to continue to perform delivery tasks normally.
[0145] If the target load of the fourth health level is a safety-critical load, such as brake lights or turn signals, the autonomous driving domain controller adjusts the vehicle task strategy, controls the unmanned logistics vehicle to terminate the delivery task, and plans for the unmanned logistics vehicle to immediately travel at a safe vehicle speed to a safe stopping point recommended by the operation and maintenance platform, waiting for operation and maintenance personnel to arrive at the safe stopping point with the necessary spare parts to replace or repair the target load.
[0146] It should be noted that if the health of the target load is less than the third health threshold (e.g., ... If the target load's health level is 40 (40 points), then the remaining lifespan of the target load is 0. The recommended maintenance time is the current moment. At this time, the target load should be stopped immediately, and an alarm should be raised to the maintenance personnel to arrange emergency maintenance.
[0147] According to the method provided in the embodiments of this application, this application also provides a computer program product, which includes: computer program code, which, when run on a computer, causes the computer to execute the unmanned logistics vehicle load health management method of any of the above method embodiments.
[0148] According to the method provided in the embodiments of this application, this application also provides a computer-readable storage medium storing program code, which, when run on a computer, causes the computer to execute the unmanned logistics vehicle load health management method of any of the above method embodiments.
[0149] like Figure 2 The diagram shows a structural schematic of an unmanned logistics vehicle according to an embodiment of this application. The unmanned logistics vehicle in this embodiment mainly includes: a vehicle body, and a chassis domain controller and an autonomous driving domain controller disposed on the vehicle body.
[0150] Among them, such as Figure 2 As shown, the chassis domain controller is connected to both the autonomous driving domain controller and the operation and maintenance platform, and is configured to execute any of the unmanned logistics vehicle load health management methods provided in the various method embodiments. The unmanned logistics vehicle load health management methods have been described in detail in the above method embodiments, and for the sake of brevity, will not be repeated here.
[0151] The autonomous driving domain controller, connected to the operations and maintenance platform, is configured to adjust the vehicle task strategy of the autonomous logistics vehicle based on the maintenance memo for each target load. Specifically, the autonomous driving domain controller obtains the health level of each target load based on the maintenance memo and adjusts the vehicle task strategy of the autonomous logistics vehicle accordingly.
[0152] In one embodiment, the autonomous driving domain controller adjusts the vehicle mission strategy of the unmanned logistics vehicle by the following steps.
[0153] ① If a target load is a safety-critical load and its health level is level 4, the autonomous driving domain controller adjusts the vehicle task strategy, controls the unmanned logistics vehicle to terminate the delivery task, and plans for the unmanned logistics vehicle to immediately travel at a safe vehicle speed to a safe stopping point recommended by the operation and maintenance platform, waiting for the operation and maintenance personnel to arrive at the safe stopping point with the necessary spare parts to replace or repair the target load.
[0154] ② If a target load is a safety-critical load with a health level of Level 3, the autonomous driving domain controller adjusts the vehicle task strategy, reducing the vehicle speed of the autonomous logistics vehicle (e.g., from 40 km / h to 25 km / h), and plans for the autonomous logistics vehicle to travel to the maintenance service point recommended by the operation and maintenance platform according to the suggested maintenance time, so that the autonomous logistics vehicle can be inspected and maintained at the maintenance service point. Simultaneously, the autonomous driving domain controller can control the autonomous logistics vehicle to reduce the use of this target load, especially avoiding its use on highways or high-load road sections. It should be noted that at this time, the health level of other safety-critical loads should be at most Level 3, Level 2, or Level 1.
[0155] ③Otherwise, the autonomous driving domain controller can ignore the health risks of each target load, keep the vehicle task strategy unchanged, and control the unmanned logistics vehicle to continue to perform the delivery task normally.
[0156] like Figure 3 The diagram illustrates the structure of a cloud-edge collaborative platform for load health management of unmanned logistics vehicles, as described in this embodiment. The cloud-edge collaborative platform for load health management of unmanned logistics vehicles in this embodiment mainly includes: multiple unmanned logistics vehicles and a cloud server.
[0157] Each driverless logistics vehicle includes: a vehicle body and a chassis domain controller and an autonomous driving domain controller installed on the vehicle body.
[0158] The chassis domain controller is connected to both the autonomous driving domain controller and the operation and maintenance platform, and is configured to execute any of the unmanned logistics vehicle load health management methods provided in the various method embodiments. The unmanned logistics vehicle load health management methods have been described in detail in the above method embodiments, and for the sake of brevity, will not be repeated here.
[0159] The autonomous driving domain controller is connected to the operation and maintenance platform and is configured to adjust the vehicle task strategy of the unmanned logistics vehicle based on the maintenance memo for each target load.
[0160] like Figure 3 As shown, the cloud server is connected to each unmanned logistics vehicle. Specifically, the cloud server is connected to the chassis domain controller of each unmanned logistics vehicle and is configured to build a health assessment model and a remaining life prediction model for each target load, and to acquire the raw operating data, health characteristic data, and health status of each target load in real time, and to periodically update each health assessment model and each remaining life prediction model.
[0161] Specifically, the cloud server can pre-build a health assessment model for each target load, providing the chassis domain controllers of each unmanned logistics vehicle with weighting coefficients for calculating the health of each target load, as shown in formula (10). , , , , , It also acquires the raw operating data, health characteristic data, and health status of each target load for each time window in real time, and iteratively optimizes the health status assessment model for each target load to update the weight coefficients.
[0162] In one embodiment, the cloud server can also pre-build an environmental compensation parameter mapping table for each autonomous logistics vehicle. This environmental compensation parameter mapping table includes environmental temperature correction factors under multiple ambient temperatures and voltage correction factors under multiple battery voltages, used to provide the chassis domain controller of each autonomous logistics vehicle with environmental correction factors when calculating the health status of each target load. Based on the real-time acquired raw operating data, health characteristic data, and health status of each target load, the impact of different operating conditions (such as temperature, vibration, and voltage fluctuations) on the lifespan of each target load is analyzed, and the environmental correction factors for each autonomous logistics vehicle are updated accordingly.
[0163] Meanwhile, the cloud server can also pre-build a remaining lifetime prediction model for each target load, so that the chassis domain controllers of each unmanned logistics vehicle can use formulas (11)-(14) to calculate the remaining available times or remaining running time of each target load. It can also acquire the original running data, health feature data and health status of each target load in each time window in real time, iteratively optimize the remaining lifetime prediction model of each target load, and update each remaining lifetime prediction model.
[0164] In one embodiment, the cloud server may also pre-calibrate and store a nominal parameter table for each unmanned logistics vehicle. The nominal parameter table stores multiple nominal values for each target load, such as nominal steady-state current, allowable surge current impact coefficient, allowable current fluctuation amplitude, allowable current offset, and rated operating life (e.g., rated number of operations or rated usage time).
[0165] The cloud server periodically updates each health assessment model, each remaining life prediction model, each nominal parameter table, and each environmental compensation parameter mapping table to the chassis domain controller of each unmanned logistics vehicle via OTA (Over-The-Air) technology, so that the chassis domain controller can call them at any time.
[0166] In one embodiment, the cloud server can also obtain maintenance memos and actual maintenance records for each target load in real time, identify common failure modes of each target load, and provide design improvement suggestions for the target load to the R&D department of the unmanned logistics vehicle, in order to reduce the failure rate of the target load and improve its service life.
[0167] In addition, the cloud server can also provide customers with visualized vehicle health reports based on the accumulated health data of the target load, enhancing customer trust.
[0168] This application achieves continuous optimization of various models or calculation parameters through cloud-edge collaboration, forming a data-driven intelligent load health management system. This ensures the accuracy of load health management for unmanned logistics vehicles, including accurate health assessment, accurate fault diagnosis, and accurate spare parts delivery. It can promptly identify early anomalies such as gradual performance degradation of target loads, intermittent poor contact, or occasional overcurrent, effectively reduce the maintenance costs of each target load, reduce the risk of unplanned downtime of unmanned logistics vehicles, and improve the operational efficiency of unmanned logistics vehicles.
[0169] Experimental data shows that the unmanned logistics vehicle, load health management method, and cloud-edge collaborative platform provided in this application can transform target load failures from "sudden failures" to "planned replacements" through early warning, reducing unplanned downtime events by more than 70%; and can shorten the target load failure location time from an average of 2.5 hours to less than 10 minutes, allowing maintenance personnel to immediately replace faulty components upon arrival without on-site troubleshooting; accurate fault diagnosis and provision of necessary spare parts for repair avoid maintenance personnel carrying incorrect spare parts, reducing ineffective work assignments by 80%; and based on health status, precise replacement of faulty target loads replaces periodic replacements. Replacing the target load can reduce spare parts consumption by 30%-50%; through health monitoring, overload, short circuit and other fault states of the target load can be detected and handled in a timely manner, avoiding cascading damage caused by long-term abnormal operation of the target load, and extending the overall service life of the target load by more than 20%; maintenance work orders are automatically generated and spare parts required for maintenance are prepared in advance, and maintenance personnel can solve problems in one visit, increasing the availability of unmanned logistics vehicles to 99.5%-99.9%; automated fault diagnosis and dispatching of the target load reduces the reliance on highly skilled maintenance technicians, and maintenance labor costs can be reduced by more than 40%.
[0170] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0171] In summary, this application provides an unmanned logistics vehicle, a load health management method, and a cloud-edge collaborative platform. Through the chassis domain controller of the unmanned logistics vehicle, it collects raw operating data from multiple target loads in real time, effectively utilizing the existing analog sampling and computing resources of the chassis domain controller to achieve non-intrusive load health monitoring. Based on multiple health assessment models, it differentiates the health status of each target load, thereby effectively improving the adaptability and accuracy of health assessment. According to each health status, it executes predictive maintenance measures and hierarchical management for each target load, automatically identifies the fault type of the target load, accurately determines the spare parts required for repair, generates maintenance memos, and reports them to the operation and maintenance platform and the autonomous driving domain controller, thus forming a closed-loop application for load health management. This maximizes the operational efficiency of the unmanned logistics vehicle while ensuring its operational safety.
[0172] Therefore, this application effectively overcomes the various shortcomings of the prior art and has high industrial application value.
[0173] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.
Claims
1. A load health management method for an unmanned logistics vehicle, applied to the chassis domain controller of the unmanned logistics vehicle, wherein the chassis domain controller is connected to the autonomous driving domain controller and the operation and maintenance platform of the unmanned logistics vehicle, characterized in that, The method includes: The system collects raw operating data of multiple target loads in the unmanned logistics vehicle in real time, and extracts features from each raw operating data to generate health feature data for each target load. Based on the pre-built health assessment model for each target load, the health of each target load is assessed according to the health characteristic data. Based on each health status, and using a pre-built remaining lifetime prediction model for each target load, the remaining lifetime of each target load is predicted, the fault type of each target load is identified, the spare parts required for maintenance of each target load are determined, a maintenance memo for each target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load and the autonomous driving domain controller can adjust the vehicle task strategy of the unmanned logistics vehicle.
2. The method for managing the load health of unmanned logistics vehicles according to claim 1, characterized in that, The raw operating data includes at least one or more combinations of: switching event monitoring data of the corresponding target load within the current time window, continuous operation monitoring data, and environmental compensation data; wherein, the switching event monitoring data includes at least one or more combinations of: the number of switching on events of the corresponding target load within the current time window, the peak value and duration of the surge current generated by each switching on event, the number of switching off events, and the back EMF drop waveform of each switching off event; the continuous operation monitoring data includes at least: the energization time of the corresponding target load within the current time window, and the steady-state current value and on-state voltage drop data at multiple sampling times; the environmental compensation data includes at least: ambient temperature and battery voltage; The health characteristic data includes at least one or more combinations of the following: surge current health characteristic data, steady-state current health characteristic data, on-state voltage drop health characteristic data, back EMF decay health characteristic data, cumulative switching count, and cumulative running time for the corresponding target load within the current time window; wherein, the surge current health characteristic data includes at least the surge current impact coefficient and the maximum surge current impact coefficient for each switching on event of the corresponding target load within the current time window; the steady-state current health characteristic data includes at least the relative deviation of steady-state current at multiple sampling moments of the corresponding target load within the current time window, and one or more combinations of the average steady-state current, peak-to-peak steady-state current, steady-state current ripple rate, and average relative deviation of steady-state current within the current time window; the on-state voltage drop health characteristic data includes at least the on-state voltage drop trend of the corresponding target load within the current time window; the back EMF decay health characteristic data includes at least the back EMF decay time constant and the maximum back EMF decay time constant for each switching off event of the corresponding target load within the current time window.
3. The method for managing the load health of unmanned logistics vehicles according to claim 2, characterized in that, All health assessment models use the same mathematical model framework, which can be expressed as: ; in, The health status of the target load is expressed as a percentage. The first health sub-item of the target load is determined based on the ratio between the average relative deviation of the steady-state current of the target load within the current time window and the allowable current deviation of the target load. The second health sub-item of the target load is determined based on the ratio between the maximum surge current impact factor of the target load within the current time window and the allowable surge current impact factor of the target load. The third health sub-item of the target load is determined based on the ratio between the peak-to-peak steady-state current of the target load within the current time window and the allowable current fluctuation range of the target load. The fourth health sub-item of the target load is determined based on the on-state voltage drop trend of the target load within the current time window; The fifth health sub-item of the target load is determined based on the changing trend of the maximum back EMF decay time constant of the target load within the current time window; The sixth health sub-item of the target load is determined as a percentage between the cumulative number of switching operations or the cumulative power-on time and the nominal lifespan of the target load. , , , , , These are the weight coefficients for the first, second, third, fourth, fifth, and sixth health sub-items, respectively. It is an environmental correction factor.
4. The method for managing the load health of unmanned logistics vehicles according to claim 1, characterized in that, After obtaining the health status of each target load, the following is also included: Based on each health level, determine the health grade of each target load; The methods for determining the health level of the target load include: The health status of the target load is compared with a preset first health status threshold, a second health status threshold, and a third health status threshold to determine the health level of the target load; wherein the first health status threshold is greater than the second health status threshold, and the second health status threshold is greater than the third health status threshold. If the health of the target load is greater than or equal to the first health threshold, the health level of the target load is determined to be the first health level, indicating that the target load is in a good state; If the health of the target load is less than the first health threshold and greater than or equal to the second health threshold, then the health level of the target load is determined to be the second health level, and the target load is indicated to be in a state of alert. If the health of the target load is less than the second health threshold and greater than or equal to the third health threshold, then the health level of the target load is determined to be the third health level, and the target load is in a warning state. If the health of the target load is less than the third health threshold, the health level of the target load is determined to be the fourth health level, indicating that the target load is in a critical state.
5. The unmanned logistics vehicle load health management method according to claim 4, characterized in that, If the health level of the target load is the second health level, then based on the pre-built remaining lifetime prediction model of the target load, the remaining lifetime of the target load is predicted, and the first maintenance memorandum of the target load is generated and reported to the operation and maintenance platform so that the operation and maintenance platform can generate a maintenance work order for the target load. If the health level of the target load is the third health level, then based on the pre-built remaining life prediction model of the target load, the remaining life of the target load is predicted, the fault type of the target load is identified, the spare parts required for the maintenance of the target load are determined, and a second maintenance memorandum of the target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller, so that the operation and maintenance platform can generate a maintenance work order for the target load and the autonomous driving domain controller can adjust the vehicle task strategy of the unmanned logistics vehicle. If the health level of the target load is level four, the fault type of the target load is identified, the spare parts required for the maintenance of the target load are determined, and a third maintenance memorandum for the target load is generated and reported to the operation and maintenance platform and the autonomous driving domain controller. This allows the operation and maintenance platform to generate a maintenance work order for the target load, and the autonomous driving domain controller to adjust the vehicle task strategy of the unmanned logistics vehicle.
6. The method for managing the load health of unmanned logistics vehicles according to claim 5, characterized in that, Methods for predicting the remaining lifespan of a target load include: If the target load has a limited number of operations, a corresponding remaining lifetime prediction model is constructed based on the cumulative number of switching operations of the target load to predict the current remaining available operations of the target load. If the health of the target load declines linearly, multiple historical health values of the target load and the cumulative running time corresponding to each historical health value are obtained. The historical health values at each cumulative running time are linearly fitted to establish a trend model of the target load health value changing with running time as its remaining lifetime prediction model, so as to predict the current remaining running time of the target load. If the health of the target load decays exponentially, multiple historical health values of the target load and the cumulative running time corresponding to each historical health value are obtained. The historical health values at each cumulative running time are then subjected to exponential fitting to establish a trend model of the target load health value changing with running time as its remaining lifetime prediction model, so as to predict the remaining running time of the target load.
7. The method for managing the load health of unmanned logistics vehicles according to claim 5, characterized in that, Methods for identifying the fault type of the target load include: If the on-state voltage drop data at multiple consecutive sampling times in the original operating data of the target load is greater than 0, but the steady-state current value is 0, then the target load is determined to have an open-circuit fault. If the target load has a steady-state current value in its original operating data that exceeds the preset short-circuit threshold of the nominal steady-state current, that is, if the steady-state current relative deviation in the health characteristic data exceeds the preset short-circuit threshold, then the target load is determined to have a short-circuit fault. If the steady-state current value at multiple sampling times in the original operating data of the target load gradually decreases or the fluctuation increases, that is, the relative deviation of the steady-state current in the health characteristic data decreases or the steady-state current ripple rate increases, then the target load is determined to have an aging fault. If the steady-state current value at one or more mutually spaced sampling times in the original operating data of the target load is 0, then the target load is determined to have an intermittent fault. If the on-state voltage drop data and steady-state current value are 0 at multiple consecutive sampling times in the original operating data of the target load, then it is determined that the load signal sampling circuit of the target load has failed.
8. The method for managing the load health of unmanned logistics vehicles according to claim 5, characterized in that, Methods for determining the spare parts required for maintenance under target load include: If the target load is replaceable, the spare parts required for maintenance shall be the standard spare parts for that target load. If the fault type of the target load is intermittent, the spare parts required for maintenance are determined to be the connection structure in the target load; If the fault type of the target load is a load signal sampling circuit fault, then the spare parts required for repair are determined to be chassis domain controller repair spare parts.
9. An unmanned logistics vehicle, characterized in that, include: The vehicle body, and the chassis domain controller and autonomous driving domain controller installed on the vehicle body; The chassis domain controller is connected to the autonomous driving domain controller and the operation and maintenance platform respectively, and is configured to execute the unmanned logistics vehicle load health management method as described in any one of claims 1 to 8. The autonomous driving domain controller, connected to the operation and maintenance platform, is configured to adjust the vehicle task strategy of the unmanned logistics vehicle based on the maintenance memo for each target load.
10. A cloud-edge collaborative platform for load health management of unmanned logistics vehicles, characterized in that, include: Multiple unmanned logistics vehicles, each unmanned logistics vehicle including: a vehicle body and a chassis domain controller and an autonomous driving domain controller disposed on the vehicle body; wherein, the chassis domain controller is connected to the autonomous driving domain controller and the operation and maintenance platform respectively, and is configured to execute the unmanned logistics vehicle load health management method as described in any one of claims 1 to 8; the autonomous driving domain controller is connected to the operation and maintenance platform and is configured to adjust the vehicle task strategy of the unmanned logistics vehicle according to the maintenance memo of each target load; The cloud server is connected to each of the autonomous logistics vehicles and is configured to build a health assessment model and a remaining life prediction model for each target load. It also acquires the raw operating data, health characteristic data and health status of each target load in real time and updates each health assessment model and each remaining life prediction model regularly.
