Low-altitude logistics distribution decision-making method and system based on multi-source data fusion

Abstract

The invention relates to the technical field of multi-source data processing, and discloses a low-altitude logistics distribution decision-making method and system based on multi-source data fusion, which are used for solving the problem of lack of a cost-benefit evaluation framework for large-scale deployment of a low-altitude logistics system in a traditional method. According to the method, firstly, a unified data basis is formed by collecting multi-source data; performing time-space alignment and weight integration on the data to generate a logistics state data set; establishing a cost model based on the data set, parameterizing equipment purchase, maintenance, network expansion and human input, and associating with a deployment scale operation strategy; utilizing the model to generate benefit indexes such as return on investment, operation efficiency and resource utilization rate; integrating the cost model and the benefit index into a closed-loop evaluation framework, and dynamically adjusting according to scene parameters; framework output is applied to distribution decision making, and management optimization of low-altitude logistics large-scale deployment is achieved through iterative adjustment and optimization of paths and resource allocation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-source data processing, in particular to a low-altitude logistics distribution decision-making method and system based on multi-source data fusion. BACKGROUND

[0002] With the rapid development of low-altitude economy, low-altitude logistics distribution, as an efficient and flexible transportation mode, has been widely used in urban e-commerce, medical supplies and emergency rescue fields; through multi-source data fusion technology such as satellite navigation, weather monitoring and sensor data integration, path planning and real-time decision-making have become a research hotspot in the industry; however, the existing low-altitude logistics distribution decision-making method and system still face significant challenges in practical application, especially the lack of cost-benefit evaluation framework for large-scale deployment, which leads to the inability to systematically quantify investment returns, operating costs and resource allocation efficiency, restricting the industrialization of low-altitude logistics; In the prior art, CN117669993A discloses a low-altitude logistics distribution path planning method based on multi-source data fusion, which fuses satellite, weather and ground sensor data, and uses Bayes inference algorithm to optimize path planning and safety warning, suitable for distribution decision-making in complex environments; although this method improves the accuracy of the path and the real-time response capability, its focus is limited to algorithm optimization for single or small-scale distribution scenarios, and it does not involve the overall cost evaluation framework for large-scale deployment, which cannot analyze comprehensive economic factors such as equipment procurement, maintenance, network expansion and manpower investment, leading to difficulty in predicting long-term investment return rate and resource utilization efficiency in actual promotion; similarly, CN116976597A discloses a low-altitude navigation and monitoring fusion service key technology based on Beidou III RDSS and logistics application demonstration, which integrates Beidou satellite data and AI prediction model to build a centimeter-level positioning and spatio-temporal fusion system, supporting safety warning for urban air traffic and unmanned aerial vehicle distribution; although this technology enhances the accuracy of data fusion and emergency response, it also ignores the economic dimension of large-scale deployment, such as infrastructure investment, energy consumption and scale effect evaluation, and cannot provide quantitative tools to optimize resource allocation, thus easily leading to cost overruns and low efficiency in multi-machine group or cross-regional operation; Although the above prior art has made progress in data fusion and path decision-making, it has not solved the core pain point of large-scale deployment of low-altitude logistics systems: the lack of a comprehensive cost-benefit evaluation framework; this makes it difficult to integrate economic indicators in the decision-making process, and cannot effectively quantify the balance between initial investment, operation and maintenance and potential income, leading to resource waste and limited scalability of the industry; therefore, a new method and system are urgently needed to address low-altitude logistics distribution decision-making based on multi-source data fusion, to build an integrated cost-benefit evaluation framework to achieve efficient and sustainable large-scale deployment. SUMMARY

[0003] In view of the deficiencies of the prior art, the present application provides a low-altitude logistics distribution decision-making method and system based on multi-source data fusion, which is used to solve the problem of lack of cost-benefit evaluation framework for large-scale deployment of low-altitude logistics system in traditional methods.

[0004] To achieve the purpose of efficient and sustainable large-scale deployment mentioned in the background art, the present application provides the following technical solutions: A low-altitude logistics distribution decision-making method based on multi-source data fusion, comprising: S1: Collecting multi-source data in the low-altitude logistics system, taking the collected multi-source data as input sources, accessing, preliminarily classifying and storing through a unified interface, and forming a unified data base; S2: Fusion processing of the collected multi-source data, aligning the data of different sources according to time stamp and spatial coordinates, and integrating into a unified logistics state data set through a preset weight distribution mechanism; S3: Based on the fused logistics state data set, a cost model for large-scale deployment scenarios is established, each cost variable is parameterized modeling, and a correlation mapping relationship is established with the deployment scale and operation strategy; S4: Constructing net cash flow based on period cost and period income sequence, discounting to get net present value and iteratively calculating internal rate of return, and under threshold constraint, efficiency and resource utilization indicators are calculated from task log; S5: Integrating the cost model and benefit indicators to form an evaluation framework, linking each element into a closed loop structure in the evaluation framework, and adjusting the framework parameters according to the parameters of different distribution scenarios; S6: Applying the evaluation framework to large-scale distribution decision-making of low-altitude logistics system, taking the framework output as decision input, and realizing distribution path and resource allocation optimization under large-scale deployment conditions through iterative adjustment.

[0005] In a preferred embodiment, multi-source data in the low-altitude logistics system is collected, the collected multi-source data is taken as input sources, and accessed, preliminarily classified and stored through a unified interface to form a unified data base, comprising: Configuring low-altitude logistics data collection scheme, taking satellite navigation position, meteorological elements, airborne sensor state and historical efficiency and failure data as multi-source input sources; Performing standardized access through a unified API interface, sequentially performing format checking, numerical range checking and integrity checking on input data, and configuring a queue buffer mechanism to queue and buffer the data arriving during peak period; Preliminarily classifying the accessed data, and generating metadata tags including data source identification, collection time stamp and priority label for each group of data; The location group, environment group, real-time group and history group are respectively stored in a partition with geographic index, time series engine, low latency storage area and relational database in the cloud storage platform, encryption and access control are implemented on sensitive fields, and data management is performed in combination with backup strategy and audit log.

[0006] In a preferred embodiment, the collected multi-source data is fused and processed, and the data from different sources is aligned according to the timestamp and spatial coordinates, including: The multi-source data is aligned in the time dimension, the timestamp of satellite navigation data is taken as the reference, the meteorological monitoring data, sensor real-time data and historical operation data are executed for synchronous matching, and the respective nearest valid records are associated with the corresponding navigation time point; After time alignment, the multi-source data is mapped in spatial coordinates, a unified geographic reference system is adopted, the positions of meteorological monitoring stations, relative azimuth information of sensors and historical path records are converted into unified global latitude and longitude coordinates, and altitude is introduced as the third dimension to interpolate height data; A weight distribution mechanism is introduced to evaluate the reliability of each data source to determine the relative weight, and the matching relationship between the weights is dynamically adjusted according to the application scenario.

[0007] In a preferred embodiment, the multi-source data is integrated into a unified logistics state data set by a preset weight distribution mechanism, including: The aligned data is integrated by layering with weights, flight state data is generated by fusing navigation and sensors, risk markers are generated by superimposing weather and associating historical records; The integrated data set is output as the input of the cost model, and the path weather equipment historical feature field is provided; The fused data is used as the unified input of the cost model construction and benefit index calculation, and is continuously processed in a predetermined order between the data layer and the model layer.

[0008] In a preferred embodiment, based on the fused logistics state data set, a cost model for large-scale deployment scenarios is established, including: The equipment procurement cost variable is defined by extracting the hardware specification requirement from the fused data set, and combining the configuration quantity and unit price; The maintenance cost variable is defined, and the repair frequency and maintenance expenditure type are extracted according to the environment data, real-time operation data and historical fault records; Based on the location data, environment data and historical data, the communication coverage density and infrastructure configuration parameters are statistically calculated, and the network expansion cost variable is defined.

[0009] In a preferred embodiment, each cost variable is parameterized and modeled, and a correlation mapping relationship is established with the deployment scale and operation strategy, including: Define the human input cost variable, and determine the monitoring post, training period, and personnel configuration parameters according to the flight status record and operation record; Convert each cost variable into period cost according to the unified aggregation granularity, calculate the period occurrence times according to the period occurrence quantity, unit price, period, and frequency, and calculate the total period cost according to the equipment procurement, maintenance, network, and manpower; Establish the corresponding relationship, associate each cost variable with the deployment scale and operation strategy, and construct the cost framework structure of the total cost and sub-cost view.

[0010] In a preferred embodiment, on the basis of the period cost and period income sequence, the net cash flow is constructed, the net present value is calculated by discounting, and the internal rate of return is iteratively calculated. Under the threshold constraint, the efficiency and resource utilization indicators are calculated from the task log, including: Construct the return on investment indicator, establish the cash flow sequence from the cost variables and historical income data, and calculate the net present value and internal rate of return; Construct the operation efficiency indicator, extract relevant variables from the data set output by the model, and calculate the unit delivery time and task success rate; Construct the resource utilization rate indicator, combine the scheduling record with the human input and equipment configuration variables in the cost model, calculate the idle time proportion of each UAV, and calculate the load balancing degree according to the dispersion degree of task allocation; Use the period cost sequence and the period income sequence to form the net cash flow sequence, calculate the net present value by discounting, iteratively calculate the internal rate of return, calculate the idle time proportion and load balancing degree under the threshold constraint, and combine the above indicators into a multi-dimensional benefit indicator set as the framework input.

[0011] In a preferred embodiment, the cost model and benefit indicators are integrated to form an evaluation framework, the elements are linked into a closed loop structure in the evaluation framework, and the framework parameters are adjusted according to the parameters of different distribution scenarios, including: Aggregate the cost variables and benefit indicators to form an evaluation element set, and establish the corresponding relationship between the equipment procurement cost, maintenance cost, network expansion cost, and human input, and the return on investment, operation efficiency, and resource utilization rate in the element set; Construct the link structure of cost and benefit in the evaluation framework, set the preset weight of the mapping path between the cost variable and the benefit indicator, and adjust the influence coefficient of the weight according to the scenario parameters; Introduce a closed-loop feedback mechanism, set a target interval for each benefit indicator, and adjust the cost variable and path configuration parameter according to the deviation between the evaluation result and the target interval; Introduce scenario parameters in the evaluation framework to dynamically adjust the weight of each evaluation indicator, and select or update the corresponding framework configuration according to the city density, flight distance, and fleet size.

[0012] In a preferred embodiment, the evaluation framework is applied to large-scale distribution decisions of low-altitude logistics systems, and the framework output is used as a decision input to achieve distribution path and resource allocation optimization under large-scale deployment conditions through iterative adjustment, including: Extracting quantitative evaluation results from the evaluation framework as decision inputs, and combining them with cost variables and scenario parameters to generate a multi-dimensional decision input set; Generating an initial distribution scheme for order demand and low-altitude transportation network, and inputting the data corresponding to the distribution scheme into the evaluation framework to calculate each benefit index and compare it with the preset target interval; Introducing an iterative adjustment mechanism in the decision application process, and performing cyclic optimization of the distribution path and resource allocation according to the evaluation deviation, performing reduction, merging, and layout adjustment of the distribution path, and order aggregation mode adjustment, and performing task migration and fleet rotation operation of resource allocation; After each round of path and resource adjustment, the scheme is re-input into the evaluation framework to obtain the index, and the iteration is terminated when the deviation is within the allowable range or the iteration round reaches the upper limit; When the order distribution, environmental parameters or cost parameters change during system operation, the cost and benefit indexes are recalculated based on the updated multi-source data, the recalculated indexes are input into the evaluation framework, and the optimization steps of path planning and resource allocation are re-executed.

[0013] On the other hand, the present application provides a low-altitude logistics distribution decision system based on multi-source data fusion, comprising: Data acquisition module: acquiring multi-source data of low-altitude logistics system, and standardizing access, classification and cloud storage through unified API interface; Data fusion module: fusion processing of collected data, alignment through timestamp and spatial coordinates, and application of weight distribution mechanism to integrate into a unified logistics state data set; Cost model construction module: based on the fusion data set, a cost model for large-scale deployment is established, equipment procurement, maintenance, network expansion and manpower investment are taken as cost variables, parameterized modeling is performed and a correlation mapping relationship is established, forming an extensible cost framework, providing a quantitative basis for benefit index generation; Benefit index generation module: generating benefit indexes using the cost model, and quantitatively calculating each index based on the relationship between the above variables to form a multi-dimensional benefit index set; Evaluation framework integration module: integrating the cost model and benefit indexes to form a closed-loop evaluation framework, linking each evaluation element through a feedback loop, and dynamically adjusting the framework parameters according to the scenario parameters to ensure that the evaluation framework adapts to different deployment requirements; Decision application module: take the output result of the evaluation framework as the input of the distribution decision, refine the path selection and resource allocation through iterative adjustment, and form a business closed loop from data collection to decision optimization.

[0014] Compared with the prior art, the present application provides a low-altitude logistics distribution decision method and system based on multi-source data fusion, which has the following beneficial effects: 1. The present application, through systematic collection and fusion of satellite navigation, weather monitoring, airborne sensors and operation history data, forms a unified and standardized data basis, providing reliable support for accurate calculation of cost models and benefit indicators. This method can accurately estimate the costs of equipment procurement, maintenance, network expansion and manpower investment through parameterized modeling and dynamic adjustment, and generate multi-dimensional benefit indicators including return on investment, operational efficiency and resource utilization, helping decision-makers optimize path planning and resource allocation in large-scale deployment environments. In addition, based on the adaptive adjustment mechanism of the evaluation framework, the system can dynamically adjust according to changes in the environment, tasks and equipment in actual operation, ensuring that the low-altitude logistics system achieves optimal cost control and resource allocation efficiency under various deployment conditions, thereby providing feasible technical support for the widespread application of low-altitude logistics and solving the problem of lack of cost-benefit evaluation framework for large-scale deployment of low-altitude logistics systems in traditional methods.

[0015] 2. The present application, by integrating satellite navigation, weather monitoring, airborne sensor data and historical operation records, forms a comprehensive real-time logistics state data basis to support dynamic optimization. Based on these data, the present application uses a combination of cost models and benefit indicators to evaluate the distribution path and resource allocation in real time, automatically adjusting the decision-making process to respond to different geographical environments, weather conditions and task requirements. The system can also continuously self-learn based on real-time equipment status and historical data, gradually improving the accuracy of path planning and resource allocation and reducing manual intervention. Through this method, the low-altitude logistics system can effectively avoid resource waste and improper scheduling, ensuring efficient execution of distribution tasks and providing an innovative solution for large-scale application of low-altitude logistics systems. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of a low-altitude logistics distribution decision method based on multi-source data fusion according to the present application is shown. Figure 2 A structure diagram of a low-altitude logistics distribution decision system based on multi-source data fusion according to the present application is shown. DETAILED DESCRIPTION

[0017] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative effort are within the scope of the present application.

[0018] Embodiment 1 Figure 1 A low-altitude logistics distribution decision-making method based on multi-source data fusion is given, including: S1: Collecting multi-source data in a low-altitude logistics system, taking the collected multi-source data as input sources, accessing, preliminarily classifying and storing through a unified interface, and forming a unified data basis; S2: Fusion processing of the collected multi-source data, alignment of data of different sources according to time stamps and spatial coordinates, and integration into a unified logistics state data set through a preset weight distribution mechanism; S3: Based on the fused logistics state data set, a cost model for large-scale deployment scenarios is established, each cost variable is parameterized modeling, and a correlation mapping relationship is established with the deployment scale and the operation strategy; S4: Generating benefit indicators using the cost model, taking the return on investment, operation efficiency and resource utilization rate as the basis for calculating the benefit indicators, and quantifying the indicators through the relationship between variables; S5: Integrating the cost model and the benefit indicators to form an evaluation framework, linking each element into a closed loop structure in the evaluation framework, and adjusting the framework parameters according to the parameters of different distribution scenarios; S6: Applying the evaluation framework to large-scale distribution decision-making of the low-altitude logistics system, taking the framework output as the decision input, and realizing the optimization of distribution path and resource allocation under large-scale deployment conditions through iterative adjustment.

[0019] S1: Collecting multi-source data in a low-altitude logistics system, taking the collected multi-source data as input sources, accessing, preliminarily classifying and storing through a unified interface, and forming a unified data basis, specifically: First, according to the low-altitude logistics scene configuration data collection scheme, the real-time position coordinate data provided by the satellite navigation system is taken as the core positioning input, wherein the position data includes longitude, latitude, altitude and speed vector, which is used to depict the motion trajectory of the unmanned aerial vehicle in the low-altitude space; at the same time, the meteorological data such as wind speed, wind direction, rainfall, temperature, humidity and atmospheric pressure collected by the meteorological monitoring station or remote sensing platform are taken as the environmental input, which is used to reflect the influence of low-altitude meteorological changes on flight safety and delivery time; in addition, the flight state and environmental perception data generated by the sensors carried by the unmanned aerial vehicle are taken as real-time feedback input, wherein the data at least include acceleration, attitude angle, battery capacity, vibration level and obstacle distance, which are used to monitor the instantaneous flight working condition and the surrounding environment; and the efficiency indicators such as average delivery time, task success rate and load utilization rate in historical operation records and fault logs such as equipment failure type, occurrence frequency and repair time are taken as historical input, which are extracted from the operation management system or log database; the multi-source data collected is comprehensively covered in the four dimensions of positioning, environment, real-time state and historical operation; On the basis of multi-source data collection, the standardized access of data is realized through a unified API interface, wherein the interface adopts a modular architecture, supports different data sources to import, parse, real-time push or batch pull according to the preset communication protocol and data format; and a data verification layer is set in the interface, which performs format checking, numerical range checking and integrity checking on the input data in turn, for checking whether the position coordinates are in the effective latitude and longitude range, whether the meteorological parameters are in the reasonable physical interval, whether the sensor data are missing or abnormal mutation, and whether the historical records are continuous on the time axis, for filtering invalid or abnormal input, wherein the numerical range checking is performed according to the range threshold preset for different data types, wherein the range threshold is determined in combination with the sensor range, physical limit parameter and historical statistical data; at the same time, a buffer mechanism is configured on the interface side, which temporarily buffers and queues the data arriving at the peak period by using the queue structure, so as to avoid the blockage caused by the writing of a large amount of data at the moment, and ensure the stable operation of the collection link; After completing the standardized access, the multi-source data entering the system is preliminarily classified: the data related to spatial position are classified into a position group, which is used to construct a trajectory sequence based on timestamp; the meteorological parameters are classified into an environment group, which is used to depict the influence of external environment on task execution; the flight state data and environmental perception data generated by the on-board sensors are classified into a real-time group, which is used to support online monitoring and anomaly detection; the operation efficiency indicators and fault logs are classified into a historical group, which is used to form the basis for long-term performance and reliability evaluation; on the other hand, the metadata tags of each group of data are uniformly generated, and include data source identification, collection timestamp and priority mark, which are used to support subsequent data retrieval, screening and fusion operation, and avoid confusion of data from different sources in the use process; Subsequently, the data of each group is stored in a structured manner in a cloud storage platform, which adopts a distributed storage architecture and allocates corresponding storage areas for different types of data: the position group data is stored in a high-availability partition supporting geographic index queries for retrieving trajectory information according to spatial range and time interval; the environment group data is stored in a storage engine adapted to time series data processing for analyzing meteorological change rules along the time axis; the real-time group data is stored in a low-latency storage area combining memory cache and persistent disk to meet online access requirements; the historical group data is stored in a relational database to record efficiency indicators and fault records through table structure and establish association between different tables using primary and foreign keys to support multi-dimensional statistical analysis; and in the data writing and access process, encryption and access control mechanisms are used for sensitive fields including location coordinates, only authorized business modules and users can decrypt and read; and regular backup and off-site redundant storage strategies are configured to improve data integrity and disaster recovery capability, while the audit log function is enabled to record operation information and timestamps of data at each link of collection, access, classification and storage, realizing traceable management of the whole data flow. Through the above collection, unified access, grouping management and structured storage of multi-source data, a reliable data foundation is built for large-scale deployment scenarios of low-altitude logistics systems, so that scattered raw data is organized into an orderly and retrievable resource library, thereby eliminating the data island and inconsistent format problems under the traditional scattered collection mode.

[0020] S2: The collected multi-source data is fused and processed, and data of different sources are aligned according to timestamps and spatial coordinates, and integrated into a unified logistics state data set through a preset weight distribution mechanism, which is specifically implemented as: First, align the data in the time dimension, taking the real-time timestamp of satellite navigation data as the reference time axis, and synchronously matching the meteorological monitoring data, sensor real-time data and historical operation data; for meteorological monitoring data, detect whether its update time falls within the preset time window centered on the navigation timestamp, if not, search for the nearest valid meteorological record forward or backward, and associate the record with the corresponding navigation point, so that each position point has corresponding weather conditions; for sensor real-time data and historical operation data, the same time window matching strategy is adopted, the sensor collection time is matched with the operation log time, the current flight state is associated with the efficiency indicators and fault records of the corresponding period, and a multi-source sequence on the unified time axis is formed, thereby avoiding time misalignment and decision bias caused by different sampling periods and update frequencies; Secondly, on the basis of time alignment, the spatial coordinate mapping is performed on various types of data to unify them to the same geographic reference system; firstly, the WGS84 coordinate system is adopted as the unified reference system, and the longitude and latitude provided by the satellite navigation are directly taken as the spatial anchor point; for the position of the meteorological monitoring station, when expressed in local coordinates or administrative grid, it is converted into the corresponding global longitude and latitude through the coordinate conversion step, wherein the coordinate conversion step is performed based on the preset coordinate conversion parameters or the conversion model provided by the geographic information system; for the relative position or distance information of the sensor real-time data relative to the unmanned aerial vehicle body, it is superimposed on the navigation coordinates to obtain the corresponding absolute position; for the path record in the historical operation data, the historical trajectory is mapped to the current reference system through backtracking processing, so that the historical path and the current running trajectory can be compared and analyzed in the same three-dimensional space; in the spatial mapping process, the altitude or flight height is taken as the third dimension, and the height information in different data sources is compared, and when the height difference exceeds the preset height difference threshold, the intermediate points are supplemented by interpolation or smoothing, wherein the preset height difference threshold can be set in combination with the unmanned aerial vehicle height control accuracy, task safety margin and historical height fluctuation statistical results, to eliminate local blank and mutation, thereby constructing a spatial view consistent in plane position and height dimension; After time and space alignment, a weight allocation mechanism is introduced, the positioning accuracy, data freshness, data integrity and historical sample size of each data source are evaluated, and the relative weight is determined accordingly; specifically, the satellite navigation data is regarded as the data source with the highest positioning accuracy, and a higher weight is given; for weather data, the weight is adjusted according to whether the interval between the update time and the current time is within the preset freshness threshold, when the data is updated in a timely manner, a higher weight is given, and when the update time interval is longer, the weight is gradually reduced, to reflect the time difference; for sensor real-time data, the weight is adjusted according to whether the data packet loss rate is lower than the preset loss rate threshold, when the loss rate is low and the data continuity is good, the weight is increased, and when the loss rate is high, the weight is reduced, to reduce the influence of noise; for operation history data, the weight is set according to whether the number of historical records exceeds the preset sample size threshold, when the sample size is sufficient, the weight is increased, and when the sample size is insufficient, the weight is appropriately reduced, to avoid small sample statistical bias; and on this basis, the relative weight relationship between navigation data, weather data, sensor data and historical data is determined, and the weight of weather data or other data sources can be adjusted according to different running scenarios, to form a dynamically configurable weighted integration basis; wherein the preset time window, the preset height difference threshold, the preset freshness threshold, the preset loss rate threshold and the preset sample size threshold can be configured in combination with the sampling period of various types of data, the sensor range, the business safety redundancy requirement and the historical running statistical characteristics, without being limited to specific numerical values; Subsequently, the aligned data is integrated hierarchically using the determined weights, and data from different sources is gradually fused into a unified logistics state dataset; first, the navigation data and sensor real-time data that have been aligned in time and space are fused to construct a core flight state subset, wherein the navigation data provides the flight path and spatial position skeleton, and the sensor data supplements dynamic details such as attitude changes, vibration levels, battery status, and obstacle distances; on this basis, meteorological data is integrated into the core subset, and environmental parameters such as wind speed, wind direction, rainfall, temperature, and humidity are mapped to each time-space point to form an environmental enhancement subset that describes the coupling relationship between position and environment; finally, the operation history data is introduced into the above data structure, and the historical efficiency indicators and fault records are associated with the current path and state, and when the current flight position or path segment overlaps with the historical fault-prone or low-efficiency area, a risk or performance label is marked on the corresponding record; through the above hierarchical integration method, a comprehensive logistics state dataset is constructed that has been integrated in four dimensions of position, environment, real-time state, and historical experience, and each data source participates in integration according to the preset weight, avoiding the deviation caused by the dominance of a single data source, and maintaining the dominant role of high-quality information in the whole; Finally, the output of the comprehensive logistics state dataset is used as the input data source for cost model and benefit index calculation, and the path features, weather risks, equipment states, and historical efficiency in the dataset are provided to the subsequent modeling step, thereby forming a smooth connection between multi-source data fusion and economic evaluation, ensuring the continuity and consistency of low-altitude logistics deployment decisions between the data layer and the model layer.

[0021] S3: Based on the integrated logistics state dataset, a cost model for large-scale deployment scenarios is established, and each cost variable is parameterized and modeled, and is associated with the deployment scale and operation strategy to form a mapping relationship, which is implemented as follows: First, the equipment procurement cost variable is defined based on the comprehensive logistics state dataset, which is used as the core component of the initial capital expenditure; by analyzing the sensor real-time data, the specification requirements of the unmanned aerial vehicle body, navigation module, and auxiliary sensors are derived, and combined with the equipment usage records in the historical operation data, the configuration quantity and reference unit price of different types of equipment are determined; accordingly, the equipment procurement cost is represented as the weighted sum of the number and unit price of each type of equipment, so that this cost can directly reflect the current deployment scale and configuration structure, avoiding static estimation that is divorced from the actual operation scale; Secondly, maintenance cost is introduced into the cost variable system to represent the repair and maintenance expenditure during the long-term operation of the system; based on the environment group and real-time group of the comprehensive logistics state dataset, factors that affect the maintenance demand are extracted, such as the frequency and duration of adverse weather calculated from meteorological data, the number of vibration level anomalies and temperature anomalies calculated from sensor data, and the corresponding fault logs in the historical group, which are divided into two types of activities, periodic maintenance and fault repair; on this basis, the maintenance cost is modeled as a function of running time, taking the cumulative running time or running age as the independent variable, so that the expected value of maintenance frequency and single maintenance cost increases moderately with time, to reflect the influence of equipment aging and fatigue accumulation on maintenance cost under large-scale deployment conditions; The network expansion cost is included in the model as an infrastructure cost variable, according to the location group of the comprehensive logistics state dataset, the spatial coverage and task density of low-altitude logistics tasks are calculated, and the deployment density requirement of communication base stations, edge nodes or relay devices is derived; combined with the parameters related to channel quality in the environment group, the redundancy capability required for data transmission link is evaluated; and the network failure rate and link interruption record in the historical operation data are used to correct the configuration requirements for network reliability and backup link; thus, the network expansion cost is divided into sub-items such as base station construction cost, link rental or bandwidth usage cost, network device upgrade cost, etc., and each sub-item is modeled according to variables such as coverage area, node density, node unit cost and business traffic, for example, the base station construction cost can be expressed as the number of nodes determined by the coverage area and node density multiplied by the single node construction cost, and the link rental or bandwidth usage cost can be expressed as a function of business traffic and bandwidth unit price, so that the network expansion cost increases significantly with the expansion of the deployment area and the increase of the task quantity, thus reflecting the changes in network investment under cross-regional and large-scale operation; Further, the human input cost is included in the cost model as an operation variable, based on the flight state, alarm events and dispatch records recorded in the comprehensive logistics state dataset, the workload demand of monitoring and dispatching is evaluated; based on the historical efficiency index analysis of task completion quality and error rate, the training demand of operation and maintenance personnel is determined; in terms of human cost, operator salary, training cost and dispatch personnel configuration are divided into several sub-variables, and a dependent relationship is established with variables such as maintenance cost and task complexity, for example, when the fault frequency statistics exceed the preset fault frequency threshold, the training frequency is increased or the on-duty personnel configuration is increased; the preset fault frequency threshold can be set in combination with the historical fault frequency distribution and safety management requirements, so that the human input cost can be dynamically adjusted with the system health status and risk level; After the definition of various cost variables, parameterize them to convert the cost structure into a calculable form; when parameterizing the cost variables, also use the evaluation period and the unified collection granularity to divide the evaluation period into consecutive periods t, and output the period t corresponding output for each cost output; construct a parameter set for each cost sub-item Solidification into the setting: Q is the scale parameter in the period T, representing the number of devices, personnel scale, network node scale, business volume, and other countable scales. The statistical caliber of the scale parameter is subject to the system metering caliber and the contract billing caliber; P is the unit cost parameter, representing the unit procurement cost, unit maintenance cost, unit rental cost, unit bandwidth cost, unit labor compensation, etc. The unit cost parameter is provided by the contract, quotation, or budget table; T is the occurrence or settlement period parameter, representing the maintenance period, settlement period, or amortization period; F is the number of occurrences per period T, and the frequency value is derived from the maintenance plan, training plan, settlement contract, or operation rule table; To implement the actual number of occurrences within the period t, define the event number formula as:

[0022] Where Δt is the period t duration and has the same unit as T, represents the ceiling; the ceiling is used to avoid underestimating the number of occurrences when collecting across periods. The rounding rule and the sources of T and F are written into the configuration, so that the number of events in any period can be directly determined by Δt, T, and F; Under the unified caliber, each cost variable is represented as quantity Q, unit price P, period T, and frequency F, and is uniformly collected according to period t. The procurement equipment cost is one-time investment, and the equipment scale parameter , unit price parameter are multiplied and added to obtain the equipment procurement cost; when depreciation or amortization is used, the one-time procurement cost is linearly amortized according to the amortization period to obtain the procurement amortization cost of each period; the maintenance cost in period t is obtained by multiplying the single maintenance cost determined by the unit price parameter and the period maintenance frequency converted according to the maintenance period and frequency, and the formula is:

[0023] Where, is the number of maintenance events in period t, i.e., the number of maintenance occurrences that should be included in the cost calculation within that period; the periodic maintenance and fault repair are calculated separately and summed to obtain the period maintenance cost; the network cost in period t is composed of expansion construction cost and running rental cost. The expansion construction cost is the new node scale parameter The product of the node unit price parameter The running lease cost is the product of the bandwidth or traffic scale parameter and the bandwidth unit price parameter ; the network fee is charged according to the settlement period , and the bill fee is proportionally allocated to each period segment and collected; wherein the manpower investment cost is multiplied by the personnel scale parameter and the per capita period salary or per capita hourly unit price parameter to obtain; the training cost is counted as an independent sub-item of the manpower investment cost, and is multiplied by the training personnel scale parameter , the single training unit price parameter and the period training times to obtain; the previous period training times are calculated according to the formula:

[0024] , wherein, is the number of maintenance events in the period t, that is, the number of maintenance occurrences that should be counted in the cost calculation in the period; t is the period number or period identifier, which is used to refer to a certain time segment after the evaluation period is divided; is the duration of the period t, and the time unit should be consistent with ; is the maintenance period, which represents the length of the period according to which the maintenance activities occur or are settled; is the training frequency based on one training period ; it represents the number of training times that occur in each training period; represents the ceiling; thus, the manpower-related cost can be determined and calculated according to the post configuration, the working hour range, the training period and the training frequency; thereby obtaining the equipment, maintenance, network and manpower sub-period costs and the total period cost collected according to the period t, which are used for subsequent net cash flow sequence construction; Subsequently, based on parameterization, a correlation mapping relationship between variables is established, linking each cost variable with deployment scale and operational strategy in a linked model. By introducing scale factors such as the number of drone clusters, average daily workload, and service area, equipment procurement costs, network expansion costs, maintenance costs, and manpower input costs are expressed as functions of scale factors. Simultaneously, based on operational strategy parameters such as communication redundancy levels, inspection frequency, and task scheduling strategies, the corresponding cost variables are adjusted. For example, equipment procurement costs are linked to network expansion costs; when the number of newly added drones exceeds a preset scale threshold, the number of base station nodes or bandwidth resources are increased accordingly. Maintenance costs are linked to manpower input costs; when the failure frequency or environmental risk indicators exceed a preset safety threshold, expected maintenance expenditures and manpower allocation levels are increased. The scale threshold and preset safety threshold can be set based on system design capacity, resource upper limit constraints, industry safety standards, and historical operational statistics, and are not limited to specific values. Through the above correlation mapping, a network structure reflecting the interaction between cost variables and their interaction with deployment scale and operational strategy is constructed, ensuring that the total cost is not merely a simple summation of individual costs but also reflects the coupling effects across variables. Based on the establishment of variable definitions, parameterized expressions, and correlation mappings, a cost model framework for large-scale deployment scenarios is constructed. The bottom layer consists of cost variables and their parameters, the middle layer describes the linkage between variables and deployment scale and operation strategy through mapping relationships, and the upper layer forms a total cost function and sub-cost views divided by dimensions such as equipment, network, maintenance, and manpower. The deployment scale parameter is used as an adjustment factor to support the smooth expansion of the model from small-scale pilot to large-scale normalized operation. The output of this cost model serves as the input basis for subsequent benefit analysis steps to further calculate various benefit indicators, so that the cost modeling results can directly serve the large-scale deployment decision of low-altitude logistics systems.

[0025] S4: Construct net cash flow based on the period cost and period revenue series, discount it to obtain net present value, and iteratively calculate the internal rate of return. Under threshold constraints, use task logs to statistically analyze efficiency and resource utilization indicators. The specific implementation is as follows: First, based on the cost model established above, the evaluation period and unified statistical granularity are determined, and cost collection, revenue recognition, task log statistics and indicator calculation are limited to the same period and the same granularity. The discount rate ρ, threshold set θ, risk weight table and load balancing mapping rules are fixed as versioned configurations and remain unchanged within the evaluation period. Subsequently, each indicator can be directly recalculated from the period cost series, period revenue series and task log. From the cash flow construction to the cash flow collection, from the cost model output to the period cost sequence, the cash outflow at least includes initial equipment investment, periodic maintenance investment, network expansion and operation investment, and human resource investment, and the time point and collection method are clearly counted: the one-time investment is counted into the cash outflow from the starting point of the evaluation period, the periodic investment is counted into the cash outflow from the occurrence period, and if there is a cross-period amortization and depreciation caliber, the amortization caliber is reflected as the cash outflow of each period according to the amortization caliber of the cost model; the operation volume related to the income is extracted from the comprehensive logistics state data set to construct the period income sequence, and the period cash inflow is determined by the number of tasks completed and the billing rules in the period, the billing rules including task type, billing unit, billing unit price and settlement caliber, and the income can be traced back and verified by the task bill one by one; the period net cash flow sequence is constructed from the period income sequence and the period cost sequence, and the period net cash flow is the cash inflow minus the cash outflow in the period, and the investment return index is uniformly inputted; In order to make the "effective completion, success, usability, rejection" judgment have executable boundaries, a threshold set θ and its source and calibration are defined in advance, and θ is used as a whole process judgment and statistical constraint; the threshold set θ should at least contain a communication threshold subset θc, an energy threshold subset θe, an airspace and compliance threshold subset θr, a safety margin threshold subset θs, and a log integrity threshold subset θd; the communication threshold defines the link quality and disconnection criterion, and contains at least one or a combination of the upper limit of continuous disconnection time, the upper limit of packet loss rate or the lower limit of link quality; the energy threshold defines the take-off and execution safety margin, and contains at least the take-off minimum energy threshold, the task process safety margin threshold and the upper limit of energy prediction error; the airspace and compliance threshold defines the boundary crossing and restricted flight constraint, and contains at least the upper limit of boundary crossing distance, the upper limit of boundary crossing duration and the restricted flight buffer zone threshold; the safety margin threshold defines the redundancy reserved on the basis of the compliance baseline to cover environmental disturbances and ensure safety; the log integrity threshold defines the missing key field placeholder and rejection rule, and contains at least the missing key field judgment condition, the allowed missing proportion or the missing duration criterion; the threshold set θ is generated based on the determination process of the rule baseline and the data calibration, the threshold θreg is set by the compliance rule base and the enterprise safety strategy, then θcal is obtained by combining the success tasks and failure task samples marked in the historical operation data, the failure omission threshold is selected as the threshold with acceptable omission rate constraint, and finally the final threshold set θ is obtained by superimposing the safety margin on the basis of the threshold; the threshold set θ records the calibration data window, the calibration criteria and the effective window with the version number, so that the evaluation result can be traced back to the source of the threshold and the judgment boundary; For the return on investment indicator, the net present value and internal rate of return are calculated using the sequence of net cash flows during the investment period and the discount rate parameter p, respectively; the discount rate p is written into the versioned configuration according to pre-defined rules, which can be the cost of capital of the enterprise or the target return on investment, or a unified discount benchmark of risk-free interest rate plus risk premium, and the source of the discount rate is reproducible; the net present value is calculated by discounting the net cash flows of each period in the evaluation period according to the discount rate p and accumulating the discounted net income, and the initial investment is the same as the single net present value; the internal rate of return is calculated by finding the discount rate that is zero, and the reproducible numerical solution method is used to solve it: first, set the search interval of the rate of return, so that the net present value function at the endpoints of the search interval has a sign change, so that the function has a root; in the interval, the interval is narrowed according to the iteration to approach the root, and the estimated rate of return is obtained, and the absolute value of the net present value in the interval is less than the pre-set error threshold or the difference between the adjacent two rates of return is less than the pre-set error threshold, and the maximum number of iterations is set to end the calculation; when the search interval does not satisfy the sign change, resulting in no root or multiple roots causing non-unique solution, output an abnormal flag according to the pre-defined rules and return the available solution strategy result, such as returning the smallest positive root or returning "no available internal rate of return" and retaining the corresponding net present value curve diagnostic information, so that the internal rate of return result has a verifiable solution path and abnormal handling boundary; Among them, the unit delivery time and the task success rate are the main ones; the unit delivery time takes the statistical object and the judgment rule as the calculation criterion; the effective task set: the task enters the execution state and enters the set, according to the log integrity threshold subset θd to check whether there are key log fields, at least there are dispatch time stamp, execution state entering time stamp, execution state exiting time stamp, track or execution segment identification, delivery confirmation receipt identification, tasks missing key fields are filled or removed according to θd, and the removal reason is recorded, so that the effective task set can be reconstructed by others according to the same rule; the unit delivery time takes the task closed loop length as the measurement index, the closed loop length starts from the time stamp when the scheduling system generates executable instructions and sends them to the fleet, and ends at the time stamp when the task arrives at the target point and completes the delivery confirmation; tasks with relay, machine replacement, charging queuing or multi-segment flight are not segmented and recalculated, and the tasks are directly compared in a closed loop; the unit delivery time period statistics take the statistical mean of the closed loop length of the effective task set as the main output, and can output the median and high quantile, the main index is a single numerical value, and the comparison between schemes is made; the task success rate takes the proportion of tasks that complete the closed loop delivery under the constraint of the threshold set θ as the measurement object, and the task success rule is: any time the task is stopped is the stop corresponding to any threshold in θc, θe, θr, θs, the trigger threshold, that is, the failed or stopped task, and the task with valid delivery confirmation receipt is successful; the task success rate is the number of successful tasks divided by the number of tasks in the execution state, and the output is a probability number from zero to one, which is corresponding to the threshold version for review; To reflect the difference of the environment without changing the main index, a risk-weighted supplementary index can be created and a determination mechanism can be provided; the risk score R can be obtained by synthesizing multi-source data, and can at least contain weather grade, airspace restriction grade, terrain complexity, communication coverage quality, task density, homogenization, and after uniform dimension, the risk score is obtained by weighted summation according to the configuration weight; the risk grading threshold is determined by the quantile point of the historical risk score distribution or the backtest result, and the grading threshold and the grading rule are set by version; the corresponding risk grade configuration weight coefficient W is solidified, the weight coefficient can be initialized by business strategy, and the weight table is finally obtained by adjusting the weight through backtest with the goal of minimizing weighted delay and failure; the weighted success rate or weighted unit distribution time is calculated by using the weight table as a supplementary output, the weighted success rate is the ratio of the weighted sum of successful tasks and total tasks respectively according to the risk weight, the supplementary output is still a probability type output and can be directly calculated in the risk score and the weight table; Idle time proportion of unmanned aerial vehicle is determined by taking idle time proportion and load balancing degree as main standards, and statistical boundary, deduction rule and mapping function are defined; the idle time proportion of unmanned aerial vehicle is described by the proportion of time not participating in task in available time, the available time is the remaining time after deducting planned maintenance period, fault downtime period and non-flyable period specified by regulations or strategy from total time of evaluation period; the planned maintenance period is determined by maintenance plan and record, the fault downtime period is determined by fault log and state machine record, and the non-flyable period is determined by airspace and operation rule table, and the deduction is constrained by corresponding criteria in threshold set θ, and abnormal label should not be expanded; the task execution time is the cumulative time length of unmanned aerial vehicle in execution state, the single machine idle proportion is the non-execution state time of single machine and the available time of single machine, and the idle time proportion of fleet is the average of idle proportion of each unmanned aerial vehicle weighted by available time, which eliminates the statistical bias caused by the difference of available time of different unmanned aerial vehicles; the load balancing degree takes whether the load distribution in fleet is uniform as the measurement unit, a unified load scale is selected and written into configuration, and one of task quantity, flight time or load quantity is selected as the main scale; the main scale load of each unmanned aerial vehicle is counted during evaluation, the relative dispersion degree of load sequence is calculated, and the relative dispersion degree is represented by the ratio of load standard deviation and average load; the unbalance degree is mapped to the balancing degree score, a monotonically decreasing mapping function is determined and disclosed, and the balancing degree is greater than the relative dispersion degree; the larger the dispersion degree is, the smaller the balancing degree is, and the mapping result is in the preset interval; when the balancing degree needs to be normalized to a fixed interval, the normalization boundary is determined by the upper and lower bounds or quantile cut-off points of the dispersion degree counted in the historical window and written into the configuration, because the abnormal value will cause the interval to stretch and the mapping can be reproduced; Finally, the index set of multi-dimensional benefits is output, including net present value, internal rate of return, unit distribution time, task success rate, UAV idle time ratio, load balancing degree, and versioned configuration; including evaluation period and granularity, threshold set θ version, risk score and weight table version, load balancing mapping rule, and recomputed from period cost, period income sequence and task log.

[0026] S5: integrate the cost model and benefit index to form an evaluation framework, link each element into a closed-loop structure in the evaluation framework, and adjust the framework parameters according to the parameters of different distribution scenarios. The specific implementation is: First, based on the cost model established in the foregoing and the calculated benefit index, the evaluation elements are aggregated. The equipment procurement cost, maintenance cost, network expansion cost and human input cost in the cost model are associated with the benefit indexes such as investment return rate, unit distribution time, task success rate, idle time ratio and load balancing degree to form an element set covering multiple dimensions of finance, efficiency and resource utilization. The equipment procurement cost is associated with the net present value and internal rate of return to measure the impact of different equipment input levels on long-term benefits. The maintenance cost is associated with the unit distribution time and task success rate to reflect the constraint effect of maintenance input on operation efficiency and reliability. The network expansion cost is associated with the unit distribution time and task success rate to reflect the support effect of network coverage and link quality on timeliness and success rate. The human input cost is associated with the idle time ratio and load balancing degree to depict the influence of personnel configuration on resource utilization and task allocation balance. Second, based on the completion of element aggregation, the link structure of the evaluation framework is constructed. Each cost variable is taken as an input node, each benefit index is taken as an output node, and multiple influence paths from cost to benefit are established. A weight or influence coefficient is preset for each path in the evaluation framework to describe the importance of a certain cost variable to the related benefit index. For example, in the scenario of focusing on safety and stability, the weight of the path between maintenance cost and task success rate and load balancing degree can be increased. In the scenario of emphasizing rapid investment recovery, the weight of the path between equipment procurement cost and net present value and internal rate of return can be increased. The weight or influence coefficient can be determined by expert scoring, simulation analysis or data fitting, etc. in combination with enterprise operation strategy, regulatory requirements and historical evaluation results, without being limited to a single setting method. Through the above organization method of input nodes, output nodes and path weights, the originally dispersed cost elements and benefit elements are integrated into a structured network-type evaluation framework, providing a basis for subsequent quantitative scoring and comprehensive sorting. A closed-loop feedback mechanism is introduced on top of the link structure, enabling the evaluation framework to have self-correcting capability. Target intervals or reference ranges are set for each benefit indicator, such as setting a minimum target interval for net present value and internal rate of return, setting a maximum acceptable time limit for unit delivery time, setting a minimum safety requirement for task success rate, setting a maximum acceptable upper limit for idle time proportion, and setting a minimum balance level for load balancing degree. When the evaluation result shows that a certain indicator deviates from its target interval, the evaluation framework generates corresponding adjustment suggestions according to the deviation direction and deviation amplitude, and feeds back to the relevant cost variables or path weights. For example, when the net present value is lower than the target interval, it is suggested to appropriately reduce the equipment procurement scale or optimize the equipment selection in the next round of scheme; when the unit delivery time exceeds the acceptable time limit, it is suggested to increase the maintenance budget or optimize the network expansion investment; when the task success rate is lower than the safety requirement, it is suggested to prioritize increasing the network expansion and maintenance investment; when the idle time proportion is higher than the upper limit or the load balancing degree is lower than the target level, it is suggested to adjust the human resource allocation or task scheduling strategy to improve resource utilization. The target interval, time limit, requirement and upper limit can be determined in combination with project planning goals, service level agreements and historical statistical data, and are not limited to specific numerical values. Through the above closed-loop feedback, the evaluation framework can not only give the current scheme evaluation result, but also provide clear adjustment direction for the next round of configuration optimization; Subsequently, the delivery scenario parameters are integrated into the evaluation framework as dynamic adjustment elements, enabling it to automatically adjust the weights and target settings according to scenario differences. Scenario parameters include but are not limited to city density, typical flight distance and fleet size. City density can be derived from geographic information and order distribution in the comprehensive logistics state data set, typical flight distance can be statistically derived from historical trajectories and planned routes, and fleet size can be determined according to the number and type of currently deployed drones. In high-density urban scenarios, the evaluation framework can appropriately increase the weight of paths related to maintenance cost and task success rate, and tighten the target intervals of task success rate and unit delivery time according to the risk level, to emphasize the importance of device reliability and operation safety. In long-distance delivery scenarios, the weight of paths related to network expansion cost and unit delivery time, task success rate can be increased to emphasize long-distance communication and continuous coverage capability. In large-scale fleet scenarios, the weight of paths related to human resource investment and idle time proportion, load balancing degree can be increased to strengthen the consideration of resource utilization and task allocation balance. Through dynamic adjustment of path weights and target intervals, the evaluation framework can smoothly switch between small-scale pilot and large-scale normal operation. Further, to adapt to the deployment needs of different construction stages, a hierarchical configuration mode is introduced in the evaluation framework, which corresponds to different weight combinations and evaluation focuses in the test stage, medium-scale operation stage and large-scale deployment stage respectively; in the test stage, the weight of the investment return type index is appropriately reduced, the weight of the operation efficiency and task success rate is increased, and the technical feasibility and operation stability are focused on; in the medium-scale operation stage, the weights of the cost and benefit paths are balanced, so that the framework focuses on cost control, efficiency performance and resource utilization; in the large-scale deployment stage, the weights of financial indicators such as net present value and internal rate of return are increased, and the adjustment strength of related feedback paths is enhanced, so as to highlight the economy and scale effect; through hierarchical configuration, the evaluation framework can cover the needs of different stages from pilot verification to full promotion, avoiding the problem of insufficient applicability of single configuration in the whole life cycle; Through the above aggregation of elements, construction of link structure, introduction of closed-loop feedback, and comprehensive application of scene parameters and stage configuration, a cost-benefit evaluation framework for large-scale deployment of low-altitude logistics system is formed; the framework integrates cost model output and multi-dimensional benefit indicators in a unified structure, which can not only give a comprehensive cost-benefit evaluation result for the current deployment scheme, but also generate targeted adjustment suggestions according to evaluation deviation and scene changes, and adapt to changes in the operation stage, thereby providing systematic and quantitative decision-making basis for deployment scheme selection, distribution path planning and resource allocation optimization.

[0027] S6: Apply the evaluation framework to large-scale distribution decision of low-altitude logistics system, use the framework output as decision input, and realize distribution path and resource allocation optimization under large-scale deployment conditions through iterative adjustment, which is implemented as follows: Based on the cost-benefit evaluation framework constructed in the foregoing, the key quantitative results used for decision are extracted from the evaluation framework, the financial indicators such as net present value and internal rate of return are used as the input of economic feasibility and expected return, the operation indicators such as unit distribution time and task success rate are used as the input of timeliness and operation reliability, and the resource indicators such as idle time proportion and load balancing degree are used as the input of resource utilization and fleet scheduling balance; at the same time, the dominant variables in the cost model such as equipment procurement cost, maintenance cost, network expansion cost and labor input cost, and the scene parameters such as city density, typical flight distance and fleet size are combined to form a multi-dimensional input set for large-scale deployment decision; After obtaining the above input set, an initial delivery scheme is generated for a given order demand set and low-altitude transportation network, where the initial scheme at least includes several delivery paths, execution frequency of each path, task allocation of different types of drones, and configuration of take-off and landing points and relay nodes, and meets basic constraints such as flight safety specifications, battery endurance, and no-fly zones; then, the cost expenditure data and operation performance data corresponding to the initial scheme are input into the evaluation framework together with the scenario parameters, and the comprehensive cost-benefit level of the scheme is calculated, and the calculated net present value, internal rate of return, unit delivery time, task success rate, idle time ratio, and load balancing degree are compared with the net present value and internal rate of return target interval, maximum acceptable unit delivery time, minimum task success rate, maximum idle time ratio, and minimum load balancing degree set in the foregoing, to obtain the deviation values of each index as the basis for subsequent adjustment; On the basis of the evaluation results, an iterative adjustment mechanism is introduced to cyclically optimize the delivery path and resource allocation scheme driven by the above deviations; when the net present value or internal rate of return is lower than the preset financial target interval, the total cost is reduced or the income expectation is improved by reducing low-value long paths, merging redundant routes, adjusting the layout of take-off and landing points, or optimizing order aggregation; when the unit delivery time exceeds the service time limit, the overall transportation time is compressed by adjusting the path topology, increasing relay nodes, increasing the execution frequency of key routes, or assigning higher flight performance models to time-sensitive tasks; when the task success rate is lower than the safety requirement, the path segments with poor weather conditions, poor communication quality, or frequent historical failures are removed, and higher stability time periods and routes are selected first; when the idle time ratio is too high or the load balancing degree is too low, tasks are migrated from low-load or long-idle drones to task-intensive areas, or the task amount and flight time are adjusted among the drone fleet to achieve more balanced resource utilization; the specific path and resource optimization process can be implemented by using heuristic search, integer programming, or simulation-based evaluation methods; After each round of adjustment, the updated path planning scheme and resource allocation scheme are input into the evaluation framework again to obtain new cost-benefit indicators, and are compared with the target interval again; when the deviations of all key indicators fall within the preset allowable range, or the iteration round reaches the preset upper limit, the iteration is terminated, the current scheme is determined as the optimized scheme under the corresponding deployment conditions, and the scheme is solidified as the execution configuration in the large-scale deployment phase; the target interval, risk threshold, and deviation allowable range can be updated according to the project planning target, service level agreement, and latest historical statistical data, or can be updated according to the project planning target, service level agreement, and latest historical statistical data, without being limited to fixed numerical values; In the actual operation of the system, when the order space-time distribution, weather pattern, cost structure or equipment state changes significantly, the comprehensive logistics state data obtained based on a new round of collection and fusion can be used to recalculate the cost model output and benefit indicators, input the latest results into the evaluation framework and start the iterative optimization process again, so that the distribution path and resource allocation are dynamically adjusted according to the changes in the operating environment; through the above application steps, the cost-benefit evaluation framework is expanded from a static evaluation tool to a decision engine driving path planning and resource scheduling, and the economic efficiency, timeliness, safety and resource utilization rate are converted into executable path and scheduling scheme in a unified structure, so as to ensure that the low-altitude logistics system realizes comprehensive optimization of controllable cost, achievable efficiency and reasonable resource allocation under large-scale deployment conditions.

[0028] The scheme of the embodiment first collects multi-source data of the low-altitude logistics system, takes satellite navigation position, weather monitoring information, flight state and environmental perception data of airborne sensors, and historical distribution efficiency and fault logs as main inputs, realizes standardized access and preliminary classification through a unified API interface, and stores in a structured manner in a cloud platform to form a unified data basis. On this basis, time and space alignment is performed on the multi-source data, the navigation time stamp is matched with the weather update time, the sensor data and operation records are mapped to a unified geographic reference system, and are combined with a preset weight to form a unified logistics state data set. Further, a cost model for large-scale deployment is constructed based on the data set, the cost variables such as equipment procurement, maintenance, network expansion and manpower investment are parameterized modeling, and are associated with deployment scale and operation strategy to calculate benefit indicators such as return on investment, operation efficiency and resource utilization rate, forming a multi-dimensional benefit indicator set. Then, the cost model and the benefit indicators are integrated into a closed-loop evaluation framework, and the framework parameters are dynamically adjusted according to different distribution scenarios, the evaluation results are called in the low-altitude logistics distribution decision-making as the basis for path planning and resource allocation, and the cost-benefit unification under large-scale deployment conditions is realized through iterative optimization.

[0029] Embodiment 2: Figure 2 A low-altitude logistics distribution decision system based on multi-source data fusion is given, which comprises: A data collection module: collects multi-source data of the low-altitude logistics system, and realizes standardized access, classification and cloud storage through a unified API interface; A data fusion module: performs fusion processing on the collected data, aligns the time stamp and space coordinates, and integrates them into a unified logistics state data set by applying a weight distribution mechanism; A cost model construction module: a cost model for large-scale deployment is established based on the fusion data set, device procurement, maintenance, network expansion and human input are taken as cost variables, parameterized modeling is carried out and a correlation mapping relationship is established, an extensible cost framework is formed, and a quantitative basis is provided for benefit index generation; A benefit index generation module: the benefit index is generated by using the cost model, and each index is quantitatively calculated based on the relationship between the above variables to form a multi-dimensional benefit index set; An evaluation framework integration module: the cost model and the benefit index are integrated to form a closed-loop evaluation framework, each evaluation element is linked through a feedback loop, and the framework parameters are dynamically adjusted according to the scene parameters to ensure that the evaluation framework adapts to different deployment requirements; A decision application module: the output result of the evaluation framework is taken as the input of the distribution decision, the path selection and resource allocation are refined through iterative adjustment, and a business closed loop from data collection to decision optimization is formed.

[0030] It should be noted that the application can be deployed on the device itself to realize embedded application, or run on PC terminal or other terminal with user interface, so as to meet various hardware environments and use requirements.

[0031] The above embodiments can be realized by software, hardware, firmware or any combination thereof, in whole or in part. When realized by software, the above embodiments can be realized in the form of computer program product in whole or in part. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on the computer, the processes or functions described in the embodiments of the application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wireless or wired direction. The wired transmission mode includes optical fiber, twisted pair, coaxial cable, etc. The wireless transmission includes infrared, microwave, etc. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center and the like containing one or more available medium collections. The available medium can be a magnetic medium (for example, floppy disk, hard disk, magnetic tape), optical medium (for example, DVD) or semiconductor medium. The semiconductor medium can be a solid state disk.

[0032] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, device and module described above can refer to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0033] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed modules can be indirect coupling or communication connection through some interfaces, devices or modules, and can be electrical, mechanical or other forms.

[0034] The modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical modules, which can be located in one place or distributed on a plurality of network modules. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0035] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically, or two or more modules can be integrated in one module.

[0036] If the functions are realized in the form of software function modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the present application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk and various program code storage media.

[0037] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0038] Finally, the above merely provides the preferred embodiments of the present application, but is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A low-altitude logistics distribution decision-making method based on multi-source data fusion, characterized in that, Comprise: S1: Collecting multi-source data in low-altitude logistics system, taking the collected multi-source data as input source, accessing, preliminary classifying and storing through unified interface to form unified data base; S2: Fusion processing of the collected multi-source data, aligning different sources of data according to time stamp and spatial coordinates, and integrating into unified logistics state data set through preset weight distribution mechanism; S3: Based on the fused logistics state data set, establishing cost model for large-scale deployment scenario, parameterizing modeling of each cost variable, and establishing correlation mapping relationship with deployment scale and operation strategy; S4: Constructing net cash flow based on period cost and period income sequence, discounting to get net present value and iteratively calculating internal rate of return, and under threshold constraint, efficiency and resource utilization indicators are calculated from task log; S5: Integrating cost model and benefit indicators to form evaluation framework, linking each element into closed loop structure in the evaluation framework, and adjusting framework parameters according to different distribution scene parameters; S6: Applying the evaluation framework to large-scale distribution decision of low-altitude logistics system, taking the framework output as decision input, and realizing optimization of distribution path and resource allocation under large-scale deployment condition through iterative adjustment. 2.The low-altitude logistics distribution decision-making method based on multi-source data fusion according to claim 1, characterized in that, Collecting multi-source data in low-altitude logistics system, taking the collected multi-source data as input source, accessing, preliminary classifying and storing through unified interface to form unified data base, comprising: Configuring low-altitude logistics data collection scheme, taking satellite navigation position, meteorological elements, airborne sensor state and historical efficiency and failure data as multi-source input source; Performing standardized access through unified API interface, sequentially performing format checking, numerical range checking and integrity checking on input data, and configuring queue buffering mechanism to queue and cache data arriving at peak period; Preliminary classification of the accessed data, and generating metadata tags including data source identification, collection time stamp and priority label for each group of data; Storing position group, environment group, real-time group and history group in cloud storage platform respectively in partition with geographic index, time series engine, low-latency storage area and relational database, implementing encryption and access control on sensitive fields, and combining backup strategy and audit log for data management. 3.The low-altitude logistics distribution decision-making method based on multi-source data fusion of claim 1, characterized in that, Fusion processing of the collected multi-source data, aligning different sources of data according to time stamp and spatial coordinates, comprising: Aligning multi-source data in time dimension, taking satellite navigation data time stamp as reference, performing synchronous matching on meteorological monitoring data, sensor real-time data and historical operation data, and retrieving the latest valid records of each and associating with corresponding navigation time point; After time alignment, spatial coordinate mapping of multi-source data is performed, adopting unified geographic reference system to convert meteorological monitoring station position, sensor relative position information and historical path record into unified global latitude and longitude coordinates, and introducing altitude as the third dimension to interpolate height data; Introducing weight distribution mechanism to evaluate the reliability of each data source to determine the relative weight, and dynamically adjusting the matching relationship between each weight according to application scenario. 4.The low-altitude logistics distribution decision-making method based on multi-source data fusion of claim 1, wherein, The unified logistics state dataset is integrated by a preset weight distribution mechanism, including: The aligned data is integrated by layering with weights, and the flight state data is generated by fusing navigation and sensors, and the risk markers are generated by superimposing weather and correlating historical records; The integrated dataset is output as the input of the cost model, and the path weather equipment historical feature field is provided; The fused data is used as the unified input for cost model construction and benefit index calculation, and is continuously processed in a predetermined order between the data layer and the model layer.

5. The low-altitude logistics distribution decision-making method based on multi-source data fusion according to claim 1, characterized in that, Based on the integrated logistics state dataset, a cost model for large-scale deployment scenarios is established, including: The equipment procurement cost variable is defined by extracting hardware specification requirements from the integrated dataset, combined with configuration quantity and unit price; The maintenance cost variable is defined, and the repair frequency and maintenance expenditure type are extracted according to environmental data, real-time operation data and historical failure records; Based on location data, environmental data and historical data statistics, the communication coverage density and infrastructure configuration parameters are defined, and the network expansion cost variable is defined. 6.The low-altitude logistics distribution decision-making method based on multi-source data fusion of claim 1, wherein, The cost variables are parameterized and modeled, and the mapping relationship with the deployment scale and operation strategy is established, including: The human input cost variable is defined, and the monitoring post, training period and personnel configuration parameters are determined according to the flight state record and operation record; Each cost variable is converted into period cost according to the unified collection granularity, and the period occurrence number, unit price, period, frequency are calculated to calculate the period occurrence number, and the total period cost is calculated by equipment procurement, maintenance, network and manpower; The cost framework structure of the total cost and the cost of the cost view is established by establishing the mapping relationship between the cost variables and the deployment scale and operation strategy.

7. The low-altitude logistics distribution decision-making method based on multi-source data fusion according to claim 1, characterized in that, On the basis of period cost and period income sequence, the net cash flow is constructed, the net present value is calculated by discounting, and the internal rate of return is iteratively calculated, and the efficiency and resource utilization indicators are calculated from the task log under threshold constraints, including: The return on investment index is constructed, the cash flow sequence is established from the cost variables and historical income data, and the net present value and internal rate of return are calculated; The operation efficiency index is constructed, the relevant variables are extracted from the dataset output by the model, and the unit distribution time and task success rate are calculated; The resource utilization index is constructed, the idle time proportion of each unmanned aerial vehicle is calculated by combining the scheduling record with the human input and device configuration variables in the cost model, and the load balancing degree is calculated according to the dispersion degree of task allocation; The period cost sequence and the period income sequence are used to form the net cash flow sequence, the net present value is calculated by discounting, the internal rate of return is iteratively calculated, the unit distribution time and task completion rate are calculated under threshold constraints, the idle time proportion is calculated, and the load balancing degree is calculated. 8.The low-altitude logistics distribution decision-making method based on multi-source data fusion of claim 1, wherein, The cost model and benefit index are integrated to form an evaluation framework, and the elements are linked into a closed loop structure in the evaluation framework, and the framework parameters are adjusted according to the parameters of different distribution scenarios, including: The cost variables and benefit indexes are aggregated to form an evaluation element set, and the corresponding relationship between the equipment procurement cost, maintenance cost, network expansion cost and human input, and the return on investment, operation efficiency and resource utilization is established in the element set; In the evaluation framework, the mapping path between cost variables and benefit indicators is set with a preset weight, and the influence coefficient of the weight on the scene parameters is adjusted; Introduce a closed-loop feedback mechanism, set a target interval for each benefit indicator, and adjust the cost variables and path configuration parameters according to the deviation between the evaluation results and the target interval; Introduce scene parameters in the evaluation framework to dynamically adjust the weights of each evaluation indicator, and select or update the corresponding framework configuration according to the city density, flight distance and fleet size. 9.The low-altitude logistics distribution decision-making method based on multi-source data fusion of claim 1, wherein, Apply the evaluation framework to large-scale distribution decisions of low-altitude logistics systems, use the framework output as the decision input, and optimize the distribution path and resource allocation under large-scale deployment conditions through iterative adjustment, including: Extract the quantitative evaluation results from the evaluation framework as decision inputs, and combine them with cost variables and scene parameters to generate a multi-dimensional decision input set; Generate an initial distribution scheme for order demand and low-altitude transportation network, and input the data corresponding to the distribution scheme into the evaluation framework to calculate each benefit indicator and compare it with the preset target interval; Introduce an iterative adjustment mechanism in the decision application process, and perform cyclic optimization of the distribution path and resource allocation according to the evaluation deviation, perform path reduction, merging and layout adjustment, and order aggregation mode adjustment, and perform task migration and fleet rotation operation on resource allocation; After each round of path and resource adjustment, input the scheme into the evaluation framework to obtain the indicators, and terminate the iteration when the deviation is within the allowable range or the iteration round reaches the upper limit; When the order distribution, environmental parameters or cost parameters change during system operation, recalculate the cost and benefit indicators based on the updated multi-source data, input the recalculated indicators into the evaluation framework, and re-execute the optimization steps of path planning and resource allocation.

10. A low-altitude logistics distribution decision system based on multi-source data fusion, used to implement the low-altitude logistics distribution decision method based on multi-source data fusion in any one of claims 1-9. Including: Data acquisition module: Collect multi-source data of low-altitude logistics system, and standardize access, classification and cloud storage through unified API interface; Data fusion module: Fuse the collected data, align them by timestamp and spatial coordinates, and integrate them into a unified logistics state data set by applying a weight allocation mechanism; Cost model construction module: Based on the fused data set, establish a cost model for large-scale deployment, take equipment procurement, maintenance, network expansion and manpower investment as cost variables, parameterize modeling and establish associated mapping relationship, form an extensible cost framework, and provide quantitative basis for benefit indicator generation; Benefit indicator generation module: Generate benefit indicators using the cost model, and quantitatively calculate each indicator based on the relationship between the above variables to form a multi-dimensional benefit indicator set; Evaluation framework integration module: Integrate the cost model and benefit indicators to form a closed-loop evaluation framework, link each evaluation element through feedback loop, and dynamically adjust the framework parameters according to the scene parameters to ensure that the evaluation framework adapts to different deployment requirements; Decision application module: Use the output of the evaluation framework as the input of the distribution decision, refine the path selection and resource allocation through iterative adjustment, and form a business closed loop from data acquisition to decision optimization.

Images